USING THE
DIGISOUND 80
MODULAR
by
W. MARSHALL & Rj
Published by:
DIGISOUND LIMITED,
14/16 QUEEN STREET
BLACKPOOL
LANGS. FY1 IPO
CONTENTS
PREFACE
SECTION
1
Introduction to the manual.
SECTION
2
Notes for general guidance.
SECTION
3
The DIGISOUND 80 modules.
SECTION
4
A step by step guide to synthesis.
4.1
Introduction
4.2
Basic keyboard patches
4.3
Addition of parallel modules
4.4
Frequency modulation
4.5
Pulse width modulation
4.6
Modulation of filters
4.7
Amplitude modulation
4.8
Synchronisation
4.9
Further dynamic control methods
4.10
The use of noise in electronic music
4.11
Applications of Sample and Hold
4.12
Ring modulation
4.13
The External Input
4.14
4.15
4.16
4.17
PREFACE
When I prepared an article for
'Electronics Today International' on using
the DIGISOUND 80 modular synthesiser
the emphasis was on more unusual
features, such as, synchronisation
techniques. It became apparent, however,
that many customers wished to start from
basics and so I asked Bill Marshall if he
would extend the original article. The
result was increasing my mere 10
diagrams to 15 in the patching section and
a proportional increase in text matter -
with more to follow.
Almost all of our customers are
experimenters who obtain a great deal of
satisfaction from exploration and from
generally doing things for themselves.
This manual caters for this capability and
the emphasis is entirely on patching the
modules together and to making full use
of the extensive control facilities
provided. At this time we have largely
avoided detailed patches aimed at
simulating specific effects or
conventional musical instruments. The
manual is, however, also in a modular
form and each section may be expanded,
or new sections added, without destroying
the pagination and general lay-out. We
will, therefore, be pleased to receive
from customers any patches of their own
origination, including specific effects and
sounds, if these are annotated with the
control settings used. We will then
publish the most useful of these as
additions to sections, or as a separate
section, acknowledging the author. We
will also be adding specific patches of our
own but the emphasis at this time must be
becoming familiar with the synthesisers
facilities and capabilities.
We have not delved into the theory of
sound. There are many published books
covering this aspect although we hesitate
to recommend titles since they may be
out of print or may not be exactly what
you require. Furthermore none are fully
comprehensive in terms of using a
synthesiser and in view of their high cost
they should be perused prior to purchase.
We recommend, therefore, that when the
opportunity arises you should take a look
around some of the major bookshops - the
music departments of Foyles in London
and Blackwells in Oxford are known to be
good sources but undoubtedly there are
many others.
Glancing through the manual, as one is
apt to do before getting down to using it,
may reveal large patches using multiple
VCO's and the like. Do not despair since
if you progress steadily through the main
using section (Section 4) in a sequential
manner you will still be able to find
plenty to do with a 'basic' synthesiser and
at the same time learn how to use the
functions provided on the modules. You
should, however, study the complex
patches since it will allow you to
determine the most rewarding way of
expanding the synthesiser. We are often
asked which modules should be purchased
but the answer must depend on what your
specific aims are. Is your interest in
using the synthesiser for music making; is
there a greater emphasis on the
fascinating electronic aspects; on
combining the synthesiser with another
instrument such as an organ or a guitar;
for specific sound generation such as a
study of percussive instrument effects;
and so on. Only you know this and your
expansion will be in the direction that
assists you in reaching your goal. With
the introduction of the 'ALPHADAC 16'
computer controller the DIGISOUND 80 is
already one of the most comprehensive
systems available and subsequent
additions to the range of products offered
will ensure that this is maintained. For
the beginner we recommend that he (or
she) starts with a very basic unit, namely,
a VCO, a low pass filter, two envelope
shapers, the dual VGA, keyboard and
power supply. One can then make a start
with this manual and build the system up
according to your aims.
We sincerely trust that this manual meets
your immediate requirements and we look
forward to its further expansion. Happy
patching.
QxjLr^\2^
1
INTRODUCTION
1.1
It is assumed that the reader will have a
knowledge of synthesiser terminology but
most of this will have been picked up
from the construction notes for the
modules and some of this is re-iterated in
Section 3 of this manual which describes
the facilities found on each module.
Likewise it would obviously help a great
deal if the reader was acquainted with
some of the basic aspects of sound but the
level required is not very high - it is a bit
like driving a car; as long as you know
how to steer it, the knobs to touch and
the 'rules' you will get along fine and you
do not need to know how the thing works,
although sometimes the latter helps!
The DIGISOUND 80 is fully described in
the general leaflet on the synthesiser and
its general specification will not be
repeated here. The main point to
remember is that it is designed with a
'plug in anything to anywhere' capability
so that no damage will occur if you
connect any of the 3.3mm sockets
together. Thus you are free to
experiment with safety and the worst that
is likely to happen is some non musical
outputs or perhaps no output at all until
you locate the wrong connection.
Another aspect we wish to stress is that
the control inputs to the modules mostly
accept to +10 volt signals and where
panel space permits an attenuator is put
on the input of the control signal. In
some instances such attenuators have by
necessity been omitted and hence we have
the 80-5 Processor module which
provides, among other things, external
attentuators and distributors. While the
control inputs can accept the full 10 volt
range if you modulate them over this
range with sharp signals, such as pulse and
sawtooth waveforms, then some control
breakthrough may occur. The same
effect will usually happen with the most
expensive electronic equipment, namely,
if you rapidly take it back and forth over
its full range then some undesirable
effects are likely to occur. In other
words, there is a difference from 'bench'
patching, where you may simply be
plugging an output of one module to the
input of another to explore its effect, and
patching for a musical effect where often
a gentle modulation gives the best results.
Likewise if you connect a 'sharp'
waveform to a VGA or VCM in a static
patch then there may be a small residual
signal at minimum control level. Again,
in practice, when one is playing the
synthesiser the waveforms are often
'softened' by filtering which immediately
reduces the level of residual signal.
Furthermore the sound output is rarely
allowed to fall to complete silence. If
you require absolute silence then it can be
obtained by various techniques. In a
modular synthesiser you have the
capability to push the modules to their
limits, and beyond, whereas in a pre-
patched synthesiser the same situation
cannot arise since it limits the users
range of control. In summary, keep a
distinction between playing the
synthesiser and playing with the
synthesiser.
A monophonic synthesiser is one of the
simplest instruments to play. A control
voltage derived from the keyboard
electronics is connected to a voltage
controlled oscillator (VCO) to determine
the pitch and the resulting waveform is
fed into a voltage controlled amplifier
(VGA) where the sound is shaped by an
envelope shaper whose control signal also
comes from the keyboard. This set-up
will produce notes directly related to the
keyboard but the resulting sound will be
very uninteresting as we are all
accustomed to listening to more complex
spectra. We have to examine ways in
which the resources of the synthesiser can
be utilised to provide useful musical and
other sound structures.
The first step is to alter the timbre of the
note so that the sound will be pleasing to
the ear and the simplest approach is
selective filtering of the harmonics in the
waveform from the VCO, Filtering is
extremely effective, and often the only
technique employed in small synthesisers,
but the use of additive and subtractive
synthesis techniques are advantageous.
For example, the sine wave from the
same WCO can be added to, or subtracted
from, the filtered sawtooth wave to vary
the levei of the fundamental frequency in
a more controlled manner. Alternatively,
other VGO's may be used to boost or cut
the partials or even introduce some
inharmonics. It has been demonstrated
that the warmth of a piano tone is due to
the fact that the upper partials are not
exact harmonic relationships of the
fundamental frequency. Naturally the
use of these more sophisticated
techniques requires a larger synthesiser
system and there is a trade-off between
striving for near perfection, which costs
money, and a sound which is acceptable.
The latter applies whether one is
concerned with imitative synthesis of a
1.2
conventional instrument
synthesis of a 'new' sound.
or
creative
Another feature to overcome is the
monotonous nature of simple synthesiser
patches, much of which arises from the
precision of electronics in contrast to the
craftsmanship, use of natural materials
and environmental effects which have a
major influence on the sound of
conventional instruments. Again there
are some very simple techniques which
are incorporated into many synthesisers
for imparting a more dynamic character
to the sound. One can, for example,
connect an envelope shaper to the voltage
controlled filter (VCF) so that the timbre
of a note is changing for at least part of
the duration of a note. Other techniques
include amplitude modulation (tremelo),
frequency modulation (vibrato), phasing or
pseudo-phasing, pitch bending and
dynamic control of pulse width (pulse
width modulation).
Reverting to the piano, a high quality
grand will have about one hundred parts
per key and the resultant sensitivity to
velocity and pressure provide the player
with a wide range of dynamic control.
These keyboard features can be simulated
electronically - at a price - but for many
applications it is unnecessary. The
important point to remember is that our
aural responses have become accustomed
to complex sound structures with a wide
dynamic range. Thus if one does not
have velocity and pressure sensitivity on
the keyboard then it is a matter of using
more cost effective resources to achieve
a dynamic character for electronically
created music.
used as building blocks for the creation of
novel sounds as well as providing a basis
for a more thorough understanding of
sound synthesis in general.
Finally, we advise the reader to work
through the manual from the beginning.
This is particularly important in Section ^
which builds up in a progressive manner
and if you attempt a patch part of the
way through without at least studying the
earlier parts then you may find it very
difficult to interpret.
The purpose of this manual is to provide
the user of the DIGISOUND 80 synthesiser
with informati9n which will enable the
best results to be obtained from the
equipment, in particular exploring some
of the aspects touched upon above. We
present a brief outline of the principal use
of each module and follow this up with
diagrams of applications that involve
several that are inter-connected or
'patched' together. The fact that one
device can control another in some way is
what allows a synthesiser to produce an
exceptionally wide range of sounds. We
do not attempt to provide a list of
specific sounds, which the user will later
be able to produce for himself, but rather
to illustrate certain standard patches and
techniques. This progresses to more
advanced developments which may be
2
NOTES FOR
GENERAL
GUIDANCE
2.1
Some of the points in this section may
appear trivial and obvious but please read
it carefully and heed the advice given,
which is intended to ailov\/ you to make
good progress through Section 4.
RECALIBRATE the modules when
necessary. This may be quite soon after
initial calibration but thereafter the
interval should be quite long. So often
one comes across synthesisers which are
not in tune, particularly amongst
constructors whose main interest is
inclined towards the electronic aspects of
music making. In the latter
circumstances the resultant output sounds
awful. If you take the time to tune it
then you will get much greater
satisfaction from the results whatever
your inclination.
BE SYSTEMATIC Devise either a flow
chart type of patch as used in Section 4 or
draw up a tabular recording system with
the outputs of the modules forming the
rows and the inputs being the columns. In
either case space should be allowed for
recording the settings of the various
control potentiometers. When a useful
sound is obtained then make a record of
the patch and describe the sound for
future reference.
UNDERSTAND THE METHOD. One may
be starting with a specific sound and
trying to improve its qualities or
modifying it to another sound. The patch
may sometimes become so complicated
that you have lost track of what is
happening. When the latter point is
reached then scrap the patch and start
again from the original point.
USE REAUSTIC METHODS. When
experimenting with the sounds of musical
instruments avoid playing notes singly.
Even if one is lacking in keyboard skill at
this time then still play a few notes at the
appropriate tempo since the end result
will be quite different and obviously more
realistic. Likewise, when creating sound
effects then ensure that the amplifier is
set to the appropriate volume - one
cannot emulate the sound of gunfire
quietly!
IMPROVE THE QUAUTY. Once a good
imitation of a conventional sound is
obtained, or a new sound created, then
one can begin to explore ways of
improving its dynamic quality. In other
words, avoid becoming too complex too
quickly but start by developing a
repertoire of sounds which may be
improved upon at a later date as you
become more familiar with the
capabilities provided in the DIGISOUND
80.
UNDERSTAND THE SOUND. Try to
work out what is happening to the sound
as it is processed. An oscilloscope is
certainly a useful aid but if one starts off
with simple patches then the ear is just as
good. After a relatively short time one
should not be in a position of finding out
what happens when, say, a high pass filter
is used for treatment. Instead, one
should be working towards the situation
where the use of a particular treatment is
a logical conclusion - this is what Section
4 is all about.
HAVE AVAILABLE an adequate number of
patchcords of varying lengths and keep
them separated in a suitable rack. If you
do this then a modular synthesiser
becomes a pleasure to use rather than an
irritation since there are few things more
frustrating than hunting through a heap of
patchcords to locate one of the correct
length. See later regarding further advice
on patchcords.
EXPAND LOGICALLY. Do not be under
the misapprehension that the modules
provided by a 'basic' DIGISOUND 80
synthesiser will allow the creation of
virtually any type of sound. On the one
hand there is much scope for additional
tonal treatment, reverberation and so on
but do not complicate matters by
obtaining these until the basic resources
have been mastered. The other aspect,
namely, increasing the number of modules
has been touched upon elsewhere. If one
starts small and then approaches synthesis
in a logical manner then future expansion
to meet specific needs should be obvious.
There is, however, much in favour of a
four 'voice' system, i.e., modules
equivalent to four times the basic patch
shown in Figure 4.1.1, with an
ALPHADAC 16 computer controller.
With such resources one may explore
complex single voice patches and at the
same time have the capability of playing
in the polyphonic mode, sequencing,
composing and using the other keyboard
routines provided which are invaluable to
both the skilled and unskilled keyboard
player.
As readers will be aware the DIGISOUND
80 is an easily expanded system but the
initial publication of the project ended
with a 'basic' system and thus high signal
levels at the final stage, the VGA, prior
2.2
to either the 80-14 modular amplifier or
an external amplifier. Such signal levels
are not ideal for external amplifiers,
mixers and the like since the input
attentuators on the latter will be near
their minimum and it is easy to overload
the external circuits and cause distortion.
As we write (Autumn 1981) there are,
however, plans to expand the DIGISOUND
80 to include other sound treatments and
some of these will be post-VCA modules.
The output signal from the latter
equipment will conform to the more usual
requirements of external equipment of
the type mentioned above. Interfacing
the synthesiser will also be dealt with in a
subsequent section (5) of this manual
entitled the 'DIGISOUND 80 ON STAGE'.
The final part of this section deals largely
with the subject of patchcords. We
recommend that all jack sockets on the
DIGISOUND 80 are grounded since this
makes it a simple matter to connect up
with external equipment of all types
which is powered from another supply.
Thus the use of screened cable is the
obvious choice for patchcords. Screened
cable also guards against picking up
extraneous signals. It is worth noting,
however, that the authors of this manual
have used unscreened cable for
interconnections within the DIGISOUND
80 without encountering any pick-up
problems. The electrical environment
will, nevertheless, vary from customer to
customer and thus we cannot vouch that
this will work for all. As a compromise
one could use unscreened wire for control
voltages, for example, the gate and
keyboard CV while retaining the screened
cords for audio signals. A particular
advantage of using single wire is the ease
of obtaining various colours which then
aid identification of routing. This is
particularly important when patching the
ALPHADAC 16 in a polyphonic situation.
In fact in this situation we recommend
making up a harness of wires to conform
to the basic patch of Figure 4.2.1 and
routing this between the modules such
that it does not interfere with the
controls. This in no way decreases the
flexibility of the system since jack plugs
may still be freely removed at either end
and, say, connected to another waveform
from the VCO. Likewise the wires for
Channel 1 within the harness may be
disconnected completely and inter-
connections made with single cords in the
usual manner when a complex patch is
required for the monophonic channel of
the ALPHADAC in the split mode. There
are pros and cons concerning the use of a
separate colour for each voice as against
different colours to identify the gate and
CV signals. In the latter case it is best
to mark the jack plugs in some way to
identify the voice channel and this may be
with a self adhesive label with the number
on or else a band of coloured tape.
Although this last part has been centred
around the ALPHADAC the same
techniques of using coloured wires and
identifying plugs are, of course,
applicable to a monophonic system. One
must, however, approach it logically since
to use a different colour for every type of
control voltage (keyboard CV, gate, LFO,
ADSR output, etc.) will greatly increase
the number of patchcords since you will
also need a variety of lengths in each
colour. The single wire for patchcords
should preferably be of a similar gauge to
the screened cable supplied but also
flexible. The type of wire often used for
test leads is ideal since it 'hangs' well, a
2.8mm diameter wire of this type is
available from Digisound Limited and
obtainable in five colours - if black is
counted as a colour!
A further aid to keeping the system tidy
is to use two leads connected to a single
jack plug as shown in Figures 2.1(a) and
2.1(b). It will generally be necessary to
slightly enlarge the hole on the top of the
plugs body using a file or a drill. The
type shown in Figure 2.1(a) is useful for
taking, say, the keyboard control voltage
to both a VCO and a VCF which may be
widely separated in distance while the
type of Figure 2.1(b) is ideal for a gate
voltage to two ADSR's since the latter
may well be en the same module, as in the
80-8D and 80-180.
FIG.2Ja
FtG.2Jb
i 1
On the subject of tidiness we are often
questioned as to the best lay-out of the
modules. The short answer is that such
an arrangement does not exist. With a
'basic' synthesiser then obviously the
2.3
modules may be arranged in such a way
that the patch of Figure 4.2.1 follows a
logical flow but as the system expands
then so arranging modules becomes more
a matter of personal preference. In
other words observe the patches you make
most often and judge for yourself whether
an alteration to lay-out will help -the
modules are easy enough to move around
at will. Another difficult question from
customers concerns advice on a proposed
system and the building of a nice cabinet
in which all the space is spoken for. It is
wrong to pre-judge the final size of a
system and any module housings should
allow for future expansion without
detracting from aesthetic appearance
which is why we have adopted a module
housing- which accommodates 12 standard
modules and which neatly interlock
together. If one is sometimes using the
synthesiser for live performance then it is
a good idea to accommodate the most
widely used modules into one or two such
cases and to house the less frequently
used modules in another case, or cases. In
this way one may perhaps avoid shipping
the complete system around. Size of
system and tidiness often go together
since in a small unit one is usually using
every module all of the time and hence
the number of connections in a limited
space looks complicated. On the other
hand when one sees pictures of Wendy
Carlos, Tomita and so on seated at their
modular synthesisers then although there
are numerous connections the overall
effect is quite pleasing. The reason for
the latter is simply that as the system
grows then one will probably only use a
proportion of the resources at any one
time. A final comment on this subject is
that although pictures of the DIGISOUND
80 show the front of the module housing
in line with the rear of the keys it is
better to step the module housing a few
inches further back (suitably propping the
module housing) since this reduces the
likelihood of cords dropping onto the keys
when one runs out of cords of the 'ideal'
length.
ADSR-1
GATE VOLTAGE 1-
ADSR-2
■H-
GATE VOLTAGE 2
FIGURE 2.2
would be difficult to set up using
conventional methods. The diode leads
may also be used to advantage when
sequencers are used for triggering
functions and it will greatly simplify
certain timing signals or gating
applications.
Another type of patchcord is one which
includes a diode connection, as shown in
Figure 2.2, but one should take the
precaution of marking which plug contains
the anode of the diode (that to ADSR 2 in
the figure). It will be evident from this
arrangement that gate voltage 1 will
cause only ADSR 1 to 'fire' but gate
voltage 2 will operate ADSR 1 and ADSR
2 together. This is potentially a useful
addition since without the diode the patch
3
THE MODULES
3.5.1
MODULE 80-5
PROCESSOR
The low output impedance and high input
innpedance of the DIGISOUND 80 modules
allow one output to drive several inputs
without overloading or introducing
appreciable errors. In order, therefore,
that a single output from a module may
have individually adjustable levels to each
of the modules that it is driving we have
placed, whenever practical, attenuators
on the inputs to modules. This
arrangement also facilitates fading in of
various effects, for example, if two
modules are being modulated from
another unit then one of the former two
may be faded in and out without affecting
the other. It was stated above that the
attenuators are placed on the inputs
whenever practical and in most cases it is
the limitation of panel space which
restricts the number that have been
included. To overcome this problem of
PROCESSOR
►) 2
o
10
e «)-...^.
LAG PROCESSOR
* 'f^ f f
10
ATTENUATOR 1
10
ATTENUATOR 2
h 2
7 -4mr~^
10
ATTENUATOR 3
10
ATTENUATOR 4
distribution we have included the 80-5
Processor module and this also includes a
few other simple functions to aid
synthesis.
Distribution of one output to four other
inputs, with any level of attenuation on
the combined output from the Processor,
may be implemented in various
combinations. For example, a single
Processor may be used to distribute one
signal to twelve other modules and with
sets of three outputs at different levels of
attenuation. For distribution purposes
alone the Processor is invaluable and at,
least one is required for every ten other
modules. Two of the distributors may
also be used as 'inverters' or as a source
of positive voltage for level shifting. To
avoid confusion the term 'inverter' will be
changed to 'SUBTRACTOR' since as
constructed the effect is to subtract the
input vdltage from +10V. The main
control signal in the DIGISOUND 80 is
based on a to +10V amplitude and thus
if, say, the output of an envelope
generator is taken via a 'subtracter' then
the attack voltage will start off at +10V
and decrease to zero instead of the
normal response of going from to +10V.
The output from the 'subtracter' has an
attenuator and thus the actual voltage
excursion may be adjusted to the range
desired. These 'subtracters' find wide
application in synthesis patching, as is
evident in the next section. It should be
noted that signals are also inverted in
phase when they pass through a
'subtracter'
Commonly the term 'inverting', especially
as applied to operational amplifiers,
means that the input voltage is inverted
in polarity, e.g., a +5V input becomes a -
5V output. There are some patches in
synthesis with the DIGISOUND 80 which
require this voltage inversion and it is
recommended that one of the 'subtracters'
is modified to an 'INVERTER'. The latter
is simply achieved by removing the 130k
resistor and 47k trimmer connected to the
inverting input on one side of the op. amp.
In addition to other uses the inverter also
allows negative DC voltages to be
obtained for offset purposes. If this
modification is made then we suggest you
mark the panel accordingly and probably
the simplest way is to prefix the 'INV by
3.5.2
an 'S' or 'P to denote 'Subtracter-Inverter'
or 'Polarity-Inverter' respectively. The
two modes are illustrated in Figure 3.5.2.
+10V
-10V—*
The other facility included in this nnodule
is the 'Lag Processor' which as the name
implies is a signal delay device akin to the
conventional portamento control on the
keyboard. Thus control signals may be
made to glide from one step to another
and provide a more subtle transition. An
example of this effect is a sawtooth
waveform from a LFO being used to
sweep a voltage controlled module. The
sharp transition as the sawtooth reaches
its peak voltage can result in an obtrusive
'plop' and the effect may be reduced, or
totally eliminated, by the Lag Processor
without detracting too much from the
intended effect.
The Lag Processor is in essence a low pass
filter with a manually adjustable cut-off
frequency in the lower frequency range.
Thus attempts to delay high frequency
signals will also result in a decrease in
amplitude of the signal. Nevertheless it
does find application as a low pass filter
especially in the treatment of white and
pink noise sources.
3.18.1
MODULE 80-18
MULTI-FUNCTION ENVELOPE GENERATOR
Few synthesists appear to recognise the
value of envelope generators other than
for the usual applications of obtaining a
sound contour when used in conjunction
with a VCA or modifying the timbre
during the course of a note when
connected to a control input of a VCF. It
should be noted that the combination of
envelope generator plus VCA is often
referred to as an envelope shaper.
Envelope generators are, however, one of
the most useful sources of control
voltages, particularly since a single gate
pulse can initiate a complex pattern of
control voltages. It is hoped that this
manual will illustrate some of the more
diverse applications.
In view of the above there is a need for
relatively low cost envelope generators to
DUAL E.G.
1 © 2
10 10
ATTACK ATTACK
DECAY DECAY
SUSTAIN SUSTAIN
RELEASE RELEASE
AUTO
3RMAL G T TtffilQ OUT
«BpEO G T TfR OUT
encourage their widespread use. At the
same time, however, quality is important
and especially the ability of the envelope
output to return to near zero at the end
of its cycle, irrespective of the settings
used. For this latter reason the
DIGISOUND 80-18 and 80-18A modules
have been introduced to replace the 80-8.
The 80-18 is a more versatile unit and has
three operating modes which are simply
selected via a single pole three-way
switch. The three modes are:-
I. NORMAL. This is the
conventional ADSR type of envelope,
illustrated in Figure 2, in which the
duration of the sustain period is
determined by the presence of a gate
voltage which in turn is equal to the
period a key is depressed.
2. NORMAL AOSR ENVELOPE
2. AUTOMATIC. In this mode a
short pulse will cause the envelope to
cycle through a complete ADR envelope
of the type illustrated in Figure 3. This
mode is useful when the module is used in
conjunction with programmable sound
generators which normally only output a
short pulse coincident with the start of a
note. It will also be found useful by less
skilled keyboard players since pressing a
key momentarily will provide a complete
envelope and one does not have to get the
sustain period timing, correct. It is also
applicable to situations where long
envelope times are set, since the user will
have both hands free to manipulate the
synthesiser while the contour is
progressing through its cycle.
^-SUSTAIN LEVEL
1.80-180 PANEL
3. AUTOMATIC ENVELOPE
3.18.2
The AUTOMATIC mode is particularly
beneficial when envelopes are being
initiated from non-keyboard sources, for
example, an LFO or the internal clock of
the 80-12 Noise Generator/Sample & Hold
module. A short pulse will now generate
a complete ADR envelope and, by
adjustment of the time constants, this
type of envelope can be made to
approximate the ADSR type, as is evident
from Figure 2. Usually these external
sources would only generate a limited AD
type of envelope.
3. DAMPED. The objective of this
mode is to more closely simulate the
piano envelope which has a sharp attack,
a brief initial decay, a long release and
fincdly a very short release as the damper
is applied to the string. This ADRR
envelope is illustrated in Figure ^. In this
mode release of the key, which is the end
of the gate pulse, causes the final release,
R 2, to occur. In other words releasing
the note has the same action as applying
the damper on a piano.
4. DAMPED ENVELOPE
The three timing functions (A, D and R)
have ranges between two milliseconds and
tej'i seconds, or more, and are
exponentially scaled. The latter results
in the most useful time ranges utilising
the highest proportion of the associated
control potentiometers. The attack
voltage rises to +10V and the sustain level
is also adjustable from zero to +10V.
The envelope generator has separate gate
and trigger inputs and their respective
jack sockets are marked G and T on the
panel. A trigger pulse is not required to
initiate any of the envelope modes but in
the NORMAL and DAMPED modes the
application of an external trigger voltage
while the gate pulse is still present will
re-start the attack cycle and thus allow
generation of multiple peaked envelopes.
The module accepts ground referenced
gate and trigger inputs within the range
of +3V to +15V. The gate input is of low
impedance and thus one should avoid
gating more than two 80-18's from an
external module, such as the 80-3 VCLFO,
whose outputs have a nominal impedance
of about ikO. The low impedance is not a
problem for the, normal gating sources,
i.e., the keyboard via 8Q-i5D2 or the
'ALPHADAC 16' since the output
impedance of these is near zero.
In the AUTOMATIC mode very high
sustain levels, about 90% or more, may
cause the ADR cycle to latch-up in some
circumstances. What will happen is the
output will stay at the maximum sustain
level. If this does occur then simply
switch to NORMAL, which will release
the cycle, and then back to AUTOMATIC
having reduced the sustain level slightly.
The time constants may be trimmed to
enable accurate matching of units in a
polyphonic system.
The module may be manually gated using
the push button marked 'MAN' and this
facility is disabled when external gate
sources are connected to the 'G' socket.
Other features of the 80-18 concern its
use with the ALPHADAC 16 operating in
the arpeggiation modes. First, the
NORMAL (ADSR) envelope must be
selected. More important is the fact
that the very short pulses generated in
the staccato mode will not re-trigger the
envelope generator. The 80-18A should
therefore be used for the monophonic
voice (split keyboard) when, or if, the
staccato effect is required.
3.18.3
MODULE 80-18A
ADSR ENVELOPE GENERATOR
The DIGISOUND 80-18A is a direct
replacement for the 80-8 module but with
improved quality in terms of control
voltage feedthrough. This improvement
ensures that the envelope returns to near
zero irrespective of the settings used for
the ADSR envelope.
The PCB and panel are identical for the
80-18 and 80-18A modules and so the
latter may subsequently be converted to a
multi-function envelope generator, if
required. A jack socket placed in the
hole used for the function switch will
improve the appearance of the module
while in its simplified form.
The gate impedance is 10k but other
characteristics are the same as the 80-18
operated in the NORMAL mode.
Reference should therefore be made to
the description of the latter module.
The 80- 18 A will be re-triggered in the
staccato mode of the ALPHADAC 16
arpeggiation routines.
4
STEP
BY
STEP
GUIDE
TO SYNTHESIS
4.1.1
4.1 INTRODUCTION
The information presented in this section
of the manual is designed to encourage
experimentation by the user and in view
of this we have omitted actual control
settings from our patching diagrams.
Thus techniques illustrated are to be used
as constructional blocks for the creation
of your own sounds or as a foundation on
which to build more complex patches.
The layout and format of this section
dealing with the applications of modular
synthesiser devices proved to be a very
difficult task owing to the fact that
similar techniques could be included in
several of the sub-sections. As a
compromise, therefore, we have tried to
create a flowing text which takes Into
account the most basic of patches while
each sub-section leads onto more
advanced methods. Because of this
approach it is necessary to work through
the sections sequentially.
Although the layout is unorthodox it is
fairly concise. Obviously the entire point
of owning a DIGISOUND 80 synthesiser is
for its great versatility and so if we were
able to list all of the patches, that could
vaguely be described as musical, then the
versatility of your synthesiser would be
very much in question. In other words, it
would be a task next to impossible. The
information is presented as a step by step
guide to the practical understanding of
electronic music techniques.
The patching configuration is as follows
and illustrated in Figure ^,1.1. All
modules will be indicated by rectangles
and the audio inputs (signals) will enter
the rectangle from the left (indicated by
'A'). Treated signals (B) will logically
follow from the right of the rectangle.
Control voltage inputs (C) will enter the
module from either the top or the bottom
while control voltage outputs (E) will
follow the same routes. Gate, trigger or
other timing signals (D) will be shown as a
dotted line if confusion is likely to arise
by using a solid line. Arrows obviously
indicate the direction of the signal or
control voltage. In many instances the
patches have been simplified by omitting
the output stages which may include some
further treatment, at the discretion of
the user, and this is signified by a bold
arrow(F). Lastly, in cases where more
than one type of input (or output) to a
module is used then the precise input (or
output) will be indicated.
As regards terminology, signals and
control voltages will often be referred to
simply as signals without distinction. The
reason being that with the DIGISOUND 80
synthesiser the two are compatible and so
only their intended use determines what
they should actually be known as.
MODULE
0^
>E
4.1.1
4.2.1
4.2 BASIC KEYBOARD PATCHES
As a starting point the DIGISOUND 80
should be patched in the same manner as
most 'mini' synthesisers, as illustrated in
Figure ^.2.1. Note that the 80-6L may be
substituted by an 80-7 in the LP'f mode or
an 80-10 may be used for each half of the
80-8D. In the arrangement shown the
frequency of the VCO is determined by
the keyboard control voltage, which is
scaled at lV/6ctave. The switch on the
VCO should be in the 'off position,
thereby disabling the octaves manual
control, while the 'fine' control should be
set to zero or tuned to A='f^0Hz. The
80-2
^
80-6L
^*!80^
VCO
"
VCF
VGA
.CI
cv
.C2
kEXP
KBD.
ADSR
^8060
ADSR
,9*r
t
'
'
M>
4.2.1
frequency scale can be transposed down
by one or two octaves or up by one, two
or three octaves by using the octave
shifter of 80-15E. Note that two
envelope generators are used, AD5R-1
and ADSR-2, one for the filter and one
for the VGA. ADSR-1 is connected to
the VCF using Control Input 2, which has
its own attentuator, and the envelope
shape allows the cut-off frequency and
hence timbre to be varied during the
course of a note. The output of the VCF
goes to the AC coupled input on one side
of the 80-9 Dual VCA (it is normal to use
the AC input for audio signals) and the
sound shaped using ADSR-2 patched to
the 'EXP' control input. Both envelopes
are gated simultaneously when a key is
pressed.
Set the patch up as foUows:-
i. VCO with square wave output, i.e.,
pulse output with manual pulse width
control (PWM) at setting 5 to produce a
50% duty cycle.
ii. ADSR-1 off by putting Control 2
potentiometer on the VCF fully anti-
clockwise.
iii. ADSR-2 set to a piano type
envelope: fast attack, high sustain and
slow release and one could start with
A=5%, D=30%, S=80% and R=90%.
iv. Resonance control on the VCF to
about 70% rotation.
Now play a sequence of notes at the
correct tempo while adjusting the
'octaves' control on the VCF. When the
resultant sound resembles a conventional
musical instrument, or something near,
then adjust the envelope of ADSR-2 to
suit. We can now begin to examine some
of the aspects of basic synthesis and this
is best achieved by conducting a series of
simple experiments. Other keyboard
sounds can be obtained by proceeding in a
systematic manner.
Modify the patch of Figure 4.2.1. as
follows:-
i. Examine the effects of a 'tracking'
filter. This is done by taking the control
voltage line from the keyboard to the
VCF Control Input 1 as well as to the
VCO, as shown in the patch. The effect
of this is such that the VCF's cut-off
frequency is directly related to the
frequency of the VCO. The actual
difference is set by the coarse (octaves)
and fine controls on the VCF and various
settings should be tried. Suppose the VCF
is set 4 octaves higher than the VCO then
since both modules are scaled to
IV/octave the harmonic content of the
notes will remain constant.
ii. Connect an attentuator (from the
80-5 Processor) to the line between the
keyboard control voltage and the VCF
Control Input 1. This will allow the
degree of tracking to be varied by the
potentiometer. In effect you are now
altering the volts/octave relationship to
the VCF which will result in a greater
amount of harmonics to be present the
higher the note played on the keyboard.
iii. Interpose a 'subtractor' (also from
the 80-5, refer to module description) in
the keyboard control voltage line to the
VCF and examine the effect at various
attentuation levels. You will find it to
be the reverse of the situation described
in (ii).
iv. Assess the influence of ADSR-1.
Special attention should be paid to the
level of the envelope, as determined by
the Control Input 2 attenuator, since high
settings may cause the filter to exceed its
4.2.2
dynamic range. In other words the
harmonic content may not change for
some of the higher notes if the filter is
also tracking. Examine different cut-off
frequencies of the VCF, with and without
the filter tracking and also with different
levels of envelope control. Also
experiment with different shapes from
the ADSR starting with simple AD
contours and progressing to full ADSR
envelopes. Normally the ADSR-1
envelope will not exceed the duration of
the ADSR-2 envelope otherwise only part
of the formers contours will be effective,
V. Carefully examine the effect of
the VCF's resonance control. As the
control is rotated clockwise a point is
reached when the filter breaks into
oscillation (not the case if you are using
an 80-7) and the point immediately before
it does so provides a very harsh but
pleasant electronic 'ring' to the sound.
This feature will be put to use later in
this section.
vi. Examine the effect of pulse width
control of the VCO. A characteristic
'phasing' sound will be heard as the pulse
width (PWM) control is turned.
vii. Repeat experiments (i) to (v) using
different waveforms from the VCO.
Generally the sawtooth waveform will be
found most useful while the sinewave
should exhibit poor response since there
are only weak harmonics (there would be
none in a pure sine wave) for the filter to
extract.
viii. Try mixtures of waveforms, using
the 80-^ Voltage Controlled Mixer (VCM)
as illustrated in Figure 4.2.2, to alter the
resultant timbre of the note going to the
VCF. The mixtures are simply obtained
by manually varying the levels of each of
the individual shapes.
using the 80-3 'subtractor' on ADSR-2.
Similarly, the 80-5 'subtractor' inverts the
phase of a signal - what happens when a
waveform (try different ones) goes direct
to the VCM and is then mixed with the
same signal after passing through the
subtractor? The object of this
programme is to familiarise yourself with
the basic keyboard patch and the
influence of all the principal controls used
in the creation of a basic sound. If you
are new to synthesis then spend plenty of
time on this section.
The above exercises will yield some
potentially useful sounds and reference
should always be made back to the
original patch in Figure 4.2.1 if in doubt
when applying the techniques to follow.
We now look towards ways of perfecting
the basic sounds and introducing greater
dynamic control over their quality. The
advantages of the modular approach in
terms of the ability to add or interpose
other modules will soon become apparent.
VCO
^ S2
VCM
A S3^
Q_S4^
hO
4.2.2
Each of the above steps should initially
commence with the patch at its original
starting point but one soon learns how to
arrive at the initial values of the controls.
Each of the steps should also be tried in
conjunction with one another and other
possibilities should be evident, such as,
4.3.1
4.3 ADDITION OF PARALLEL MODULES
VCO
1
VCO
— 1
VCF
-,
vco
S2^
VCF
C>
r
1— ^
VCM
vco
^
VCM
L ^
VCF
D
vco
J ' — -
vco
VCM -
L
vco
VCO
4.3.1
One of the most important ways of adding
to the resources of a synthesiser is to add
more VCO's to enrich the sound. Two
VCO's controlled by the same keyboard
voltage could be taken to Signal Inputs 1
and 2 on the VCF and tuned together in
unison, near unison, or to any interval
such that the beat frequency is not
obtrusive. There are various others ways
we can connect additional VCO's and a
few of these are illustrated in Figures,
^.3.1, ^.3.2 and ^.3.3. As you can see,
the versatility of the modular approach to
synthesis is becoming apparent and the
patching possibilities very numerous.
Each of the examples demonstrate often
subtle but noticeably different effects.
The concept can be extended a stage
further by duplication of modules. In the
next example (Figure ^.3.^) the same
ADSR's are used to control both 'voices'
and variation in the envelope amounts and
the filter cut-off frequencies can be
adjusted, as previously described, to alter
the timbre of each voice. Note that to
realise this example the Processor module
is being used as a 'patchcord splitter' and
for such a basic function it is not shown in
the block diagram.
4.3.2
4.3.3
VCO
I
VCF
. t
VCA
1
•i
1
-|
1
111*.
1
CI
C2
EXP
VCM
,._
vco
2
[ 1
VCF
2
Li
VCA
2
'
' 1
F"
C2
EXP
Cv
KBD
ADSR
1
ADSR
2
GATE '
i
4.3.4
]D
Other techniques, similar to those
outlined previously, can also be added to
this patch. As an aid we offer a few
experimental guidelines:-
i. ADSR-1
VCA-2.
to control VCF-1 and
ii. One voice, or only the VCF or the
VCA, is inverted in phase with respect to
the other, i.e., a 'subtractor' sub-module
is put into the control line to the
module(s) in question.
iii. Filter tracking.
iv. Bring in unfiltered sound from the
VCO's directly to the other channels of
the VCM.
V. Taking another waveform from
each of the VCO's to the opposite filter as
illustrated in Figure ^.3.5.
vco
VCF-L
L
"1
vco
1 »
VCF-L
10
4.3.5
Adding a third filter, Figure 4.3.6, further
increases the possible effects obtainable
from a single patch. Note that in these
last two patches we have not shown the
ADSR's, etc. and they should be extended
as in Figure 4.3.4.
vco
— *
VCF-L
-*
VCA
l^
VCM
VCF
BorP
4-
vco
t;
VCF-L
-*
VCA
^
^ VCM Q
4.3.6
Again combinations of earlier examples
can be applied here and a staggering
range of possibilities may be effected.
None of the patches are, however,
modulated in any way other than from the
envelope generators controlling the
filters. There are three main types of
modulation we require to study before we
can make an in depth study of sound
synthesis. . The three are frequency
modulation (FM), pulse width modulation
(PWM) and amplitude modulation (AM)
and are discussed in the following
sections.
4.4.1
4.4 FREQUENCY MODULATION
To demonstrate frequency modulation
start with the basic patch shown in Figure
^.2.1 and connect the triangular output
(the +10V output will give more meaning
to your settings) from an 80-3 VCLFO to
the Control Input 2 of the VCO.
Experiment with different frequencies of
modulation, different modulating
waveforms and different settings of the
Control 2 attenuator.
It will be noticed that increasing the
amount of modulation with Control 2
attenuator determines the 'swing' or
amount of deviation from the set
frequency. Also experiment with
temporarily putting the switch(es) on the
VCO(s) to the ON position to bring in the
use of the coarse control. Variation of
the coarse control will alter the
frequency which is being modulated. An
obvious point you may say but now
progress by altering the VCO frequency in
such a way that the deviation from VCO
frequency to peak modulation frequency
is one octave, or more, by careful
adjustment of Control 2 on the VCO. A
square wave VCLFO output may yield the
best results. Now tune in a second VCO
in the same way, perhaps even using
different VCLFO's for each VCO and
modulated at different rates. A very
pleasing rich sound will normally result
and one must continue to experiment in
this way to fully utilise the possibilities
that exist.
The effective frequency range of the
VCLFO may be altered by application of a
negative voltage to the Control Input 1.
The 'subtractor' is a source of positive
voltage (refer to SO-5 module description)
and when passed through the 'inverter'
becomes a negative one. This may then
be used to manually extend the VCLFO's
frequency, see Figure ^.^.1.
VCLFO
VCLFO
. tci
jj"
SUBT.
-
INV.
H
VCLFO
4.4.1
4.4.2
Another interesting technique to explore
is modulating the modulator as shown in
Figure ^.^.2. The concept may obviously
be extended to modulating the modulator
which modulates another modulator!
Remember it is both the subtleness and
richness of these effects which makes the
modular synthesiser such a powerful
instrument. Pay special attention also to
the use of high modulating frequency
settings since very complex and unusual
effects can be obtained owing to the
frequency harmonics of the modulating
source being 'superimposed' on the main
audio signal, especially at high levels of
modulation. Figure 4.^.3 shows a simple
patch that can yield some unusual sounds
using two VCLFO's. Note once more that
the large arrow indicates that the patch
would normally be extended in the usual
way with the keyboard control voltage
going to VCO Control Input 1, the output
going to a VCF and onto a VCA and both
of the latter modules controlled by
ADSR's gated from the keyboard. Figure
^AA shows a simple method of adding a
pleasing richness to the sound from two or
more VCO's, with or without the ranges of
modulation being present as discussed
above. In this case gentle modulation
works best which is achieved by
adjustment of the Control 2 attenuator.
VCLFO
____
VCO
3:
ID
VCO
]C>
VCLFO
VCO
]D
VCLFO - '
VCO
10
VCO
10
44-3
4.4.4
In the above examples the modulation has
been applied to Control Input 2 which is
scaled at one volt per octave or
proportions thereof according to the
setting of Control 2 potentiometer. The
80-2 VCO's also provide for linear
modulation and the input and associated
attenuator is marked 'FM'. Using this
input will result in a linear change in
frequency with applied voltage. This*
input has some special uses. First by
using an envelope generator (ADSR plus
VCA) the modulating waveform may be
shaped. This is shown in Figure ^.4.5. A
more usual application is the arrangement
of Figure ^AA and if the VCO's are set to
different intervals then a type of
chorusing effect will be obtained since
the tracking of the oscillators has been
affected. The FM input should be
4A2
explored but unless specifically stated
frequency modulation (FM) in the rest of
this text will refer to modulation of the
exponential input (CI or C2).
vco
t
FM
VCLFO
— »•
VCA
LIN
ADSR
GATE
■
■VC074)
4-4.5
4>*.6
Cross modulation techniques are now
quite popular but the effects are quite
complex and the following patches should
be studied with care so as to determine
what is happening. Cross modulation is
achieved when we patch the output of
module B to the input of module A and
then the output of module A to the input
of module B. One such patch is shown in
Figure ^.4.6. Surprisingly the output
waveforms are stable since the modules
settle into a complex equilibrium. Figure
^.4.7 is an extension of the method and
note the use of the FM input here. When
these patches are controlled by the
keyboard (to the VCO's) the resultant
sounds, at different pitches, are often of
totally different timbres and sometimes
do not seem to have any relation to each
other.
vco
1
VCLFO
VCO
2
3t
^
VCLFO
VCO
■>
^^~
C2
VCO
i
SUBT.
C2
4.4.7
4A8
Frequency modulation, however, need not
confine itself to automatic repetition of
sounds, such as vibrato which is frequency
modulation at around 7Hz, or involve
itself in complex cross modulation
patches. Some very simple, yet
effective, ways of modulating the
frequency of a VCO may be achieved with
minimum effort. Take, for example,
Figure 4.4.S. Instead of both VCO
frequencies rising one is going up as the
other goes down. At slow modulation
rates the effect is more noticeable,
especially if the modulation and the
VCO's are 'tuned'. Some novel variations
of this can be implemented when a
keyboard is introduced by applying a
subtracted version of the keyboard
control voltage to only one of the VCO's
leaving the other in the normal mode.
Bring in cross modulation to the patch and
you have a rather unpredictable keyboard!
Frequency modulation may be
accomplished in many ways, one simply
looks around the synthesiser for sources
of control voltages which may be useful
for such purposes, e.g., the External Input
or the Sample & Hold as will be discussed
later. One of the most useful control
sources is the ADSR envelope shaper and
only a few synthesists seem to realise its
application beyond the conventional VCA
and VCF shaping techniques. Patch an
ADSR output to the Control Input 2 of the
VCO and gate the patch, as usual, from
the keyboard. The frequency will follow
the ADSR contour with the range of
frequency being adjusted by the Control 2
attenuator on the VCO. The classic, now
cliche, drum synthesiser sound can be
imitated by using an ADSR patched in this
way. . A fast attack and a moderate
decay with the sustain level set to zero
will result in a 'one-shot' type of contour
which is similar to the drum synthesiser
sound. Since the sustain level is zero the
position of the Release time control is
irrelevant. By applying this AD one-shot
to a VCO, VCF and VCA the patch is
complete. Experiment with the
resonance control. As the filter goes
into oscillation, or approaching same, at
high Q settings (high resonance) a
secondary tone will merge with the
original tone. The frequency of the
secondary tone, which is a sine wave, may
be adjusted using the filter frequency
controls. In fact the filter alone can
provide the sound when it is used in the
high Q mode. Experiment with different
ADSR contours, inverted and subtracted
contours, and the mixing of two, or more,
contours. The re-trigger or delay modes
on the 80-10 VCEG can be used to
advantage in these situations. Figure
4.^.9 illustrates the latter and the same
combination may be used in many other
applications of envelope contours. In
fact this dual peak type of contour is
often more realistic when simulating the
sounds of specific instruments. Although
J GATE
I ADSR TRIGGER
80-10 TRIGGER
4.4.9 COMBINING ENVELOPES
4.4.3
the example is a simple one, that may be
approximated using the re-triggering
facility on the 80-10 VCEG alone, the use
of two or more ADSR contours, with the
addition of delay and re-triggering when
required, can be used to generate unusual
and complex contours. To obtain a
better feeling for the contour shape you
are generating it is best to set them up
using a VCO rather than a VGA since
small changes in frequency are easier to
detect than small changes in amplitude.
At some stage you should go back to the
earlier sections and evaluate these
contours in some of the more useful patch
settings that you have evolved.
4.5.1
4.5 PULSE WIDTH MODULATION
Pulse width modulation (PWM) is a well
known technique but surprisingly many
synthesisers, including very costly units,
do not provide the facility. The
DIGISOUND 80 VCO and VCLFO are
provided with both manual and external
pulse width control which allows adjust-
ment of the pulse width over its full duty
cycle. At zero setting (PWM Control at
indicates zero percent duty cycle) no
sound will actually be produced from the
pulse output. As the PWM control is
turned clockwise the pulse wave gets
wider and becomes a square wave (50%
duty cycle) at setting 5. Further rotation
widens the pulse until at full rotation the
output is virtually a DC voltage. As
mentioned in the section concerned with
the basic keyboard patch a very pleasant
phasing sound will be produced as the
pulse width control is turned. The
flexibility of the facility is, however,
more fully realised by automatic control
of pulse width. Patch the DC triangle
wave from a VCLFO, set to about 2 or 3
cycles per second, to the PWM socket of
the VCO. The PWM Control poten-
tiometer is now disabled. The phasing
effect will now be automatic but there
will also be times when there is no sound
output, albeit momentarily. The reason is
that since the triangle is ramping between
and 10 volts the pulse width output is
cut off at these two extremes, as
described for the manual adjustment.
This can be used to advantage, especially
at high PWM frequencies when a form of
amplitude modulation will be super-
imposed on the output. The effect is
quite different from other types of
amplitude modulation discussed in the
next section. For more conventional use
of PWM, however, it will be necessary to
lift the 'floor' of the modulating
waveform above zero to provide a smooth
and uninterrupted pulse width modulated
output. This may be accomplished using a
positive voltage derived from the 80-5
'subtracter' sub-module and adding it to
the modulating waveform of the VCLFO
using the voltage controlled mixer. The
arrangement allows the level of
modulation to be precisely adjusted over
any useful range of values. The patch
may then be extended by taking the
output from the VCM and gently pulse
width modulating three or more VCO's,
possibly adding attenuation (from the 80-
5) to control the level of PWM to each
VCO. The patch is shown in Figure 4.5.1
and it may be used to achieve a very
warm choral effect.
4.S.1
4A2
An unusual type of modulation occurs
with the patch of Figure 4.5.2. We have
assumed the use of two VCLFO's although
in many cases the effects will apply if
only one is used. Alternatively, if only
one VCLFO is used then the method may
be explored using the patch shown in
Figure 4.5.3. For these experiments use
the VCO's in the free-run mode, that is,
without using the keyboard and with the
octave switch on the VCO's to the ON
position to allow manual adjustment of
their frequency. The VCO's are cross
modulated, one VCO using the linear FM
input and the other using the exponential
input via Control Input 2. Both are
modulated by sawtooth waves. Certain
rules should be applied to obtain the best
effects:-
i. Pulse width modulate the VCO's
one at a time and use high frequency
settings of the VCLFO's such that the
output seems to suddenly lock and
produce a smoother tone than that
obtained at other settings. The VCO's
should initially be tuned to within a few
hertz of each other and also set to fairly
low audio frequencies. Note that no
positive offset is applied to the PWM
control from the VCLFO's since such a
stepr is unnecessary at high frequency
settings.
ii. Fade in the frequency modulation
to each VCO, one at a time, until similar
locking effects are obtained.
iii. Next adjust the cut-off frequency
of the VCF to obtain the best sound.
There are several other aspects of the
patch which should be observed:-
4.5^
i. A very powerful sound results
which carries a great deal of weight at
the bass end of the spectrum.
ii. The resultant waveform is of an
extremely complex nature.
iii. The success of the patch largely
depends on the settings of level for both
the pulse width and frequency modulation.
iv. Each VCO should produce a
different timbre from the other and by
mixing at the filter stage yet another
timbre is developed.
The patch may now be experimented with
in the following manner:-
i. Exchange linear to exponential
inputs and vice versa on each VCO thus
altering the cross modulation.
ii. Only use the linear control input
(FM) for cross modulation of frequency.
iii. Only use the exponential input (C2)
for cross modulation.
iv. Experiment with different
waveforms from the VCLFO's which are
being used for PWM.
V. Experiment with different
waveforms for cross modulating and also
different VCO outputs to the filter.
vi. After carrying out the above steps,
and combinations thereof, choose a few of
the most interesting and examine the
influence of the frequency control
settings on both the VCLFO's and the
VCO's.
vii. Add a positive voltage offset
both PWM inputs, as described earlier.
viii. Cross modulate the VCLFO's.
to
There are numerous other ways in which
the basic patch of Figure 4.5.2 may be
configured. Figures 4.5.3 to 4.5.6 show
some variations that should be tried and
notes made of the effects obtained.
Figure 4.5.6 illustrates a method for using
the VCM module to control the heart of
the effect and their variations which is
achieved by alteration of the input levels
and panning controls. Another variation
is the substitution of a VCM for the VCF
in the initial patch of Figure 4.5.2 and
also taking the VCLFO outputs to the
final audio mix.
Study these patches carefully and try
adding some of your own. There are
several reasons for introducing such
complex patches at this stage: 1. As an
introduction to complex patching. It is a
difficult task and one which requires a
great deal of practice if the 'spaghetti'
networks are to produce interesting or
even musical sounds. 2. They teach that
precise setting of controls is necessary in
the creation of useful effects and as a
result one learns to be patient, which is
essential for good creative synthesis.
3. They teach method and progression.
One soon learns the best way of quickly
arriving at initial settings of controls and
which modules to bring into the patch and
in which order. 4. They encourage
experimentation and after a while a
similarity in some of the effects will be
observed. At this stage, therefore, one
should be beginning to form a sound
picture in the mind as to how the various
degrees of complexity can be produced
with the minimum of effort and how the
addition of another module will affect the
result.
VCO
5
VCLFO
VCO n,J
PWMnica
LAG
VCF-L
VCO
VCLFO H
-C2l
i »wmT
VCLFO
VCO
PWM A
VCF
VCLFO
VCLFO n
VCO
D
TFM
VCO
[pwmICZ-
VCO
L IPWM*
p
VCM
VCLFO
APWM
lT fff
VCM
VCLFO
VCO
4,
^^Ivco tr
L3
VCF
4.5.3
4.5.4
4.5.5
4.5.6
4.6.1
4.6 MODULATION OF FILTERS
Many of the characteristic sounds of the
synthesiser are largely due to filtering. In
the following discussion we will make use
of the four types of filter offered in the
80-6 series, namely, low pass, high pass,
band pass and phase shift. In most cases
the 80-7 state variable filter may be
substituted for the 80-6 filters, for
example, LP^ on the latter gives the same
response as the 80-6L. Also if using the
80-7 filter one should examine the
difference obtained with the two pole
outputs. In the following examples the
filter will be designated by VCF followed
by suffixes -L, -H, -B and -P denoting low
pass, high pass, band pass and phase shift
respectively.
The most common type of filter used in
synthesisers is the low pass type and when
a filter is unspecified, i.e., merely written
as VCF, then the filter type in nearly all
synthesiser texts will be a low pass
version.
Refer back to the basic patch of Figure
4,2.1. As shown the most widely used
method of modulating the filter is by the
introduction of an envelope contour from
an ADSR unit which varies the cut-off
frequency of the filter as the note
progresses and finally decays. Intro-
duction of the keyboard control voltage,
or a portion of it, produces a sound with
more dynamic character and is a
technique which should be widely used.
The keyboard control voltage is normally
injected into Control Input 1, which has
no attenuator, and is available for filter
tracking purposes. If the full degree of
tracking is not required then the keyboard
voltage may be attenuated by using
Control Input 2 and its associated
potentiometer or if this input is being
used for another purpose an 80-5
attenuator to Control Input 1 may be
used.
In the first series of experiments,
evaluate the various filter types in the
basic keyboard patch and study the
effects of using different levels of
envelope and resonance. Also introduce
filter tracking as discussed above.
Refer to the patch in Figure 4.6.1. The
addition of a high pass filter to the basic
keyboard patch, in series with the low
VCF-L
VCF-H
VCF-L
p-» ***! 1. 1 1^
]c> -u=rK>
4.6.1
4A2
pass filter, is the only modification
required. A little thought here will
reveal that if the cut-off frequencies do
not overlap then the arrangement creates
a band pass filter where the width of the
band can be made very wide. Experiment
with controlling these two filters as if
they were one by tracking them and also
modulating them from a single envelope
generator. Unusual timbres can result if
the envelopes are introduced in different
proportions to each filter.
Figure 4.6.2 shows the same two filter
types in parallel, which effectively
creates a band reject (notch) filter.
Repeat the series of experiments applied
to Figure 4.6.1 and pay particular
attention to the cut-off frequencies of
the filters and also to the resonance
control, all of which will influence the
sharpness of the notch.
The use of the filter combinations
described above offer greater synthesis
possibilities than the use of a band pass or
notch filter alone. The reason being the
ability to vary the width of the pass or
reject bands.
It will be obvious from earlier discussions
that the VCLFO may be used to modulate
the cut-off frequencies of VCF's and for
this purpose the waveform should be
connected to Control Input 2 so that the
attenuator may be used. Try different
waveforms and also the effects of
differing amounts of modulation to each
filter in the examples of Figures 4.6.1 and
4.6.2 when configured within the basic
patch. Some very interesting and useful
results should emerge.
The band pass filter may be used to
enhance some intermediate frequencies of
a waveform. Most conventional musical
instruments have certain natural resonant
frequencies and to simulate this effect
these frequencies may be boosted using
band pass filters, usually more than one
being required for realistic simulation.
The effect is obtained by setting the band
pass filter to a frequency coincident with
4.6^
the resonant frequency of the particular
instrument being simulated and combining
the resultant signal with the unfiltered
signal in a VCM, or at the input of
another filter, and subjecting the output
to normal processing via a VGA etc.
Figure ^.6.3 shows the use of a VCM for
the patch. Adjustment of the levels of
signals at the mixing stage will determine
the degree to which the resonant
frequency is being boosted and the
controls allow complete adjustment from
resonant frequency only (band pass
output) to no resonant frequency
(untreated signal). Increasing the Q
factor of the filter by turning the
resonance control clockwise will
effectively reduce the bandwidth of the
frequencies enhanced as well as providing
additional boost. Generally moderate to
high levels of Q are most effective but
any tendency to oscillation should be
avoided in this situation.
VCF-B
VCM
n
4J6.3
4J6.4
As inferred earlier, the latter technique
may be extended by using two or more
band pass filters in parallel to produce
two or more resonant peaks which can be
made to track the keyboard and produce a
type of voltage controlled resonator.
A similar type of patch may be used for
the removal, or attenuation, of some
unwanted intermediate harmonics. One
simply removes the band pass filter in
Figure ^.6.3 and substitutes it with the
band reject patch of Figure ^.6.2 or by
using the 80-7 filter in the notch mode.
Again we can experiment further with
these patches and most usefully with the
resonant frequency boosting type. Start
by examining the effects obtained from
envelope shapers, keyboard voltages and
VGLFO's to shift the resonant peak. Next
take portions of the envelope waveform,
keyboard voltage or VGLFO waveform
through an attenuator or 'subtractor' and
thence to the appropriate control input
for external voltage control of the signal
levels entering the VGM. The result of
the latter experiment will be dealt with in
more detail later in this manual but one
should' see, or rather hear, that many
more possibilities of dynamic control are
available as soon as we begin to modulate
the mixing levels, hence the timbre, of
the treated sound. As a primer on
dynamic control methods, apply the
techniques of VGLFO, keyboard voltage
or envelope output to the mixing levels of
the patches illustrated elsewhere in this
manual, in particular to those shown in
Figures 4.3.3, 4.3.4, 4.5.1, 4.5.5 and 4.5.6.
These techniques, however, are really a
form of amplitude modulation which is
dealt with in greater detail in the next
section.
The 80-6P phase shift filter has two deep
notches in its output signal. The phasing
effect will be obtained by modulating the
filter. It is better to have more control
over the proportion of treated and
untreated signals and to take the outputs
to a stereo pair in which only one side has
the phasing effect, see Figure 4.6.4. An
80-6P filter in series with the 80-7 in the
notch mode will produce an even greater
effect. Additional experiments with this
combination are applying different
modulation rates to each filter using the
ADSR or VGLFO waveform outputs.
As mentioned in the description of the
modules, the 80-6 filters can be made to
resonate, i.e., to break into oscillation at
high Q (resonance) settings. We can
therefore add another technique to our
steadily increasing repertoire of sounds by
making use of the filters ability to
oscillate. Figure 4.6.5 shows a patch
which does not require a signal to be
supplied to the VGF. The 80-6L works
best in this application. The effect being
produced is to shock the filter into
oscillation and the resultant sine wave is
shaped by the ADSR connected to the
VGA in the usual manner. The frequency
of the sinewave output is determined by
the keyboard control voltage plus the
setting of the coarse and fine controls on
the filter. The resonance control should
be set fully clockwise. If we now mix in a
little of the envelope voltage via the
Gontrol Input 2 on the filter we again
obtain the cliche drum synthesiser sound.
VCF
VCA
■c.
t
KBD
-1
ADSR
e*
TE
i
ID
4£^
All of the filters have provision for
external control of resonance. Look back
over your experiments and choose those
where the setting of the resonance
control had a useful effect. These should
now be repeated and the resonance
controlled using voltage controlled
sources such as VGLFO, keyboard voltage
or ADSR output. If the filter has
provision for simultaneous manual and
external control of resonance then
4.6.3
remember that the manual control will
determine the initial Q of the filter and
to begin with this should be set to zero.
Furthermore the modulating voltage
should have an attenuator in-line and the
80-6 filters may be arranged so that the
manual control is disabled by insertion of
a jack plug and the potentiometer for the
manual control may then be used for the
external control voltage.
4.7.1
4.7 AMPLITUDE MODULATION
Amplitude modulation is quite simply the
modulation of the level of a particular
signal. In a small synthesiser the tremeio
effect, which is amplitude modulation at
about 7H2, is usually confined to applying
modulation from a low frequency
oscillator to a control input on the VGA -
although many lack even this basic
facility. This is a sad omission since we
will demonstrate that amplitude
modulation is a very powerful tool which
may be used in a large number of
situations. The 80-9 Dual VGA has a to
100% amplitude modulation control which
therefore allows a wide range of control.
The usual method of amplitude control
applied to the VGA does have one draw-
back, namely, that the whole sound
envelope is modulated. The latter effect
is far from a natural tremeio which does
not commence immediately nor is
constant in frequency for the duration of
the note. We will limit ourselves to
discussion of the various applications of
amplitude modulation and to outlining a
few methods of transforming the usual
boring and unrealistic synthesiser tremeio
effect into a more interesting sound.
Some of the effects are quite subtle but
when used in a musical piece it is often
the small changes which provide
character and interest to the music.
The first variation in amplitude
modulation arises when the signal from a
VGA (one half of an 80-9) is amplitude
modulated by a (VG)LFO and the output
taken to another VGA (the other half of
the 80-9) where it is amplitude modulated
by another LFO at a slightly different
frequency. Envelope shaping of the input
signal is carried out in the first VGA in
the normal manner via the EXP input. If
necessary, the two modulating
frequencies may be derived from a single
LFO if the Lag Processor (80-5) is used in
the control path to the second VGA, Sine
or triangle waves are normally used for
AM purposes but experiment with other
waveforms, their combinations, and with
different modulating frequencies.
Remember that the 80-9 requires a to
10 volt input to the AM control.
Another interesting effect is split phase
tremeio in which the same signal source
goes to separate VGA's (both halves of the
80-9). The VGLFO is connected direct to
VGA-i while for VGA-2 the LFO
waveform is phase inverted by the 80-5
'subtractor' sub-module. The two outputs
are taken to separate channels of a stereo
amplifier. The arrangement is shown in
Figure ^.7.1. Envelope shaping can be
applied to both VGA's in the normal
manner or by using another VGA on the
signal input before it is split.
In the essence the 80-^ VGM module is a
quad VGA and in the previous section we
saw how it could be used to alter the
timbre of a sound by additive synthesis
techniques. This approach will now be
extended. If a number of signals are
being mixed in the VGM then amplitude
modulation of one or more of the signals
can be obtained via the appropriate
voltage control input. This will result in
the creation of new timbres depending on
the patch used. Envelope shapers may be
used for the modulating voltage or indeed
a proportion of the keyboard control
voltage which will then give a degree of
mixing in direct relationship to the notes
being played on the keyboard.
VGA
!><]'
VCLFO
VCM
». SUBT.
CiT C2^
VCLFolL SUBT. L
4.7.1
4.7.2
The following examples of amplitude
modulation may be applied to any
previous patch shown in the manual where
the VGM has been used. Figure ^.7.2 is a
variation of the split phase tremeio patch
since the mixing levels of signals 1 and 2
will rise and fall in opposite phase. Signal
3 (and 4 if used) is unaffected in this
situation. The keyboard voltage could be
substituted for the VGLFO in this patch,
or used on the other two channels.
Figures ^^.7.3 and i^JA carry this a stage
further by using envelope contours
initiated from the keyboard. Subtractors,
lag processors and attenuators may be
used to advantage in these patches which
are in addition to the attenuators on the
input signals which allow for different
initial signal levels. The signal input
paths to the VGM are omitted for reasons
of clarity.
4.7.2
VCM
AOSR.
K80. OR
VCLFO
VCM
i. .L U
SUBT.
SUBT.
D ^
LAG
Aosn,
KB0.OR
VCLFO
ATT.
SUBT.
ATT.
4.7.3
4.7.4
Figure ^.7.5 shows a patch which uses
three modulating voltages simultaneously.
In practice you will require attentuators
in the control lines since the maximum
control input to the VCM is +10 volts and
combination of VCLFO with the keyboard
control voltage or the ADSR will exceed
this level. Reverting to Figure 4.7.4
the response to control voltages and
hence the signal outputs from the
four channels will be as follows :-
MODULATING CJ,
VOLTAGE
C2
C3 C4
Low (OV) Low High Mid Low
High (lOV) High Low Low Mid
The versatility of this technique should
now be apparent and the variations are
almost limitless. Having given the
guidelines you should experiment with
further combinations.
VCM
_S <
JtJ
:.|r4l_
T
SUBT.
SUBT.
]
1
... I
KBD.
VCLFO U
ADSR
-*■ VCF
VCA
VCLFO
fl—* ADSR
4.73
4.7.6
The 80-10 VCEG allows all four of the A,
D, S and R functions to be controlled by
external voltages. By applying a VCLFO
to the external sustain control via one of
the attenuators on the 80-10 it is possible
to amplitude modulate the signal only
during the sustain phase of the envelope.
This is very effective in playing modes
which use a long sustain period since the
tremelo only occurs during the sustain.
Variations of this can be made by
•referring to Figure ^.4.9 which would
allow the amplitude modulation to occur,
say, only during the period of the second
envelope. Additionally, although not a
part of amplitude modulation, is the
ability to alter the time constants of the
80-10 VCEG by applying external control
voltages to the appropriate control input.
The most obvious application is to make,
say, the attack or release time
proportional to the keyboard control
voltage although some very strange
envelopes can be realised by using
different VCLFO modulating frequencies
at each of the time constant and sustain
inputs.
Percussive effects are also a form of
amplitude modulation which use a
repeating envelope contour in place of a
simple waveform. In Figure 4.7.6 the
ADSR unit is gated automatically from
the pulse output of a VCLFO. The pulse
width in this application determines the
gate 'on' time, hence the time before
release occurs. Short percussive
envelopes work best, such as, simple AD
contours without sustain. The patch may
be extended by applying a proportion of
the VCLFO or ADSR voltages to the filter
control inputs. Furthermore, if the 80-10
is used in this technique then the time
constants may be subjected to
modulation, as indicated above.
In all the applications where a VCA is
used with an ADSR the resultant envelope
may be partly buried, thereby reducing
the output level, by applying part of the
keyboard voltage to the VGA's AM or
Linear control inputs. Another result of
burying the envelope is to reduce the
release time. The latter can be put to
good use since although one can still have
a long release time it will sound sharper
since only the fastest part of the
exponential decay is heard. Another
method of obtaining automatic control of
loudness is to use the voltage controlled
panning facility of the VCM. If only one
output is taken from the VCM then the
amplitude of the mixed signals is
proportional to a voltage applied to the
VC pan input. Similarly, if the other
output of the VCM is used with the same
panning control voltage (the pan control
must be either fully clockwise or anti-
clockwise) then the amplitude of the
mixed signals will be inversely
proportional to the modulating voltage.
By using both outputs from the VCM the
result is, of course, automatic panning
where the signal moves from the left
output to the right output and vice versa
in time with the modulating voltage. This
is an effective technique but one which
should be used sparingly. To provide
variations of the technique we could
patch the keyboard voltage to the VC pan
input so that the sound moves, say, from
left to right as the keyboard is ascended.
This is an excellent way to increase the
'sound stage' over which the synthesiser
4.7.3
will perform in a stereo set-up and the
technique should be thoroughly explored.
A similar effect, although a more
pronounced one, will occur if an ADSR
contour is employed. What we are aiming
to do is point out some of the many
possibilities. We do not suppose that you
will ever wish to play a conventional tune
in which the sound is panned around by an
envelope shaper but for an original
composition or effect it may give just the
results you are seeking. A further
development of panning is a situation in
which the signals 'appeared' and
'disappeared' as the mix is panned from
one channel to the other. In Figure kJJ
the input signal 2 can be made to fade out
and in as the envelope voltage dies away
and the sound has travelled from left to
right and back again. Removing the
'subtractor' from the patch causes the
reverse of the situation regarding signal
2. Unusual situations arise if the
modulating voltage is derived from a
VCLFO run at fairly high frequencies.
VCM
i
C2
'
\
v<
ADSR
SUBT.
1 t
VCM
MODIFIER
-0
4.7.7
4.7.8
Since the VCM has two outputs, left and
right, and the signal mix can be panned
between them we can use this facility as
a means of providing voltage control or
timbre. Each output may be taken to a
separate modifier chain and processed
individually and by voltage control of the
panning the resultant signal will
demonstrate a more dramatic change in
timbre than has been described so far,
especially when the envelope generator is
used as the control voltage. The sound
can then be made to totally change in
quality as it is shaped. Figure ^.7.8
shows the basic arrangement of this
voltage controlled timbre modulator and
the MODIFIER is any means of signal
modification from simple filtering
onwards.
vco
1
of
■51*
EUL
VCO loj
2
S4
VCO p-"
3
VCM
VCLFO
ADSR
-S^
VCF-L
ID
SUBT.
T
:GATE
VCF-B
r
VCEG
KBD.
VCLFO — i
IC-
4.7.9
As indicated, the VCM is a particularly
useful module since it may be used in
many different roles. Since, however,
amplitude modulation of the outputs
provide the panning effect described it is
the entire mix which is subjected to the
overall panning effect. Referring back
to Figure 4.7.9 we can see that VCO-1
sawtooth output is by-passing the VCM to
remain fixed in the sound stage. This
approach of taking one, or more, signals
which by-pass some treatment stages is
an extremely important one to recognise.
Panning, of a type, may also be
accomplished with the patch of Figure
^.7.10. In essence this is a split phase
tremelo patch but using two different
sound chains instead of the same signal
feeding each VCA. The VCLFO may be
replaced by an envelope generator or, to
lesser effect, by the keyboard control
voltage. This opposite mode from panning
is known as SEQUE and its use should be
apparent.
•*> VCA
SUBT.
VCO
1
vco
2
- VCLFO
,,LINOfi
-^ VCA
VCM
cijot
SUBT.
3"
ADSR
VCLFO
The patch of Figure ^.7.9 combines a few
of the suggestions made this far. As an
exercise work out what is happening in
the patch as soon as a key is pressed and
over the duration of the envelopes. Do
this first without reference to the
synthesiser and then check your
conclusions with an audible demonstra-
tion. Adding final envelope shaping to
VCA's added to both outputs will yield a
fully configured patch.
4.7.10 4.7.11
All of the examples may, of course, be
expanded by combining them with earlier
techniques. They serve to illustrate the
variety of sounds that can only be
obtained with a modular synthesiser and
which will prove beneficial to the user
once the basic techniques have been
mastered. Figure 4.7.11 is another patch
incorporating some of the techniques
discussed above.
4.8.1
4.8 SYNCHRONISATION
When two waveforms of very nearly the
same frequency, or multiples thereof, are
mixed together a periodical inter-
amplitude modulation of the sound occurs.
This effect is known as 'beating' and can
serve to increase the richness of the
sound or can, on the other hand, be a
totally undesirable component of the
mixed waveforms.
Synchronisation of two or more
oscillators, set to ratios to produce
complex waveforms, is required when the
beat frequency has to be eliminated from
the sound. The effect of synchronising
oscillators is to lock the harmonic
relationships between two, or more,
together so that the combination sounds
more like one oscillator with a very
complex waveform. This effect must
remain when the frequencies of each
oscillator are altered simultaneously, for
example, from the keyboard. The
technique may, however, also be used in
its own right to produce some pleasing, or
unusual timbral effects. It will be
evident that synchronisation applies to
the use of two or more VCO's but the
technique may be explored by using a
VCLFO/VCO combination especially when
the LFO is set to higher frequencies.
The effect of HARD SYNCHRONISATION
is illustrated in Figure 4,8.1. A positive
going synchronisation pulse will cause the
triangle waveform to reverse direction
only during the rising portion of the
triangle whereas a negative going
synchronisation pulse will cause reversal
only during the falling portion. The
effect of the synchronisation pulses on
some other waveforms is also illustrated
and since the sine wave is derived from
the triangle wave in the VCO it will also
influence the sine wave shape in a
complex manner.
-^ — ' I ' I
4-B.1
POSITIVE HARD SYNCHRONISATION
may be implemented as illustrated in
Figure 4.8.2. VCO-1 (or a VCLFO for
experimental purposes) is referred to as
the master oscillator and the pulse output
from this is connected to an attenuator on
the 80-5 Processor module. The four
outputs from the attenuator allow up to
four oscillators, known as slave
oscillators, to be synchronised by
connection to the +HS input on the VCO's.
VCO
1
n
r
r
\^
VCO
2
VCO
3
VCO
4
4.8^
Additional oscillators may be
synchronised from the same master by
using one of the attenuators outputs
connected to the input of another
attenuator to increase the number of
distribution outlets. Although Figure
4.8.1 shows the synchronisation pulses as
spikes the VCO's respond to the positive
going edge of the pulse and so the duty
cycle of the actual pulse is not critical.
The 80-5 attenuator is used solely for
distribution purposes and although
synchronisation may occur at lower
settings the potentiometer should be set
fully clockwise at the start. Some
experiments with synchronisation should
now be conducted:-
i. Examine various frequency ratios
of master and slave oscillators and note
the influence that synchronisation has oh
the harmonic content of the output. It
may be best to start with one slave VCO
and then add others, if you have them.
The frequency of the slave VCO's should
be higher than that of the master VCO
otherwise the slave(s) will simply follow
the frequency of the master.
ii. Repeat step (i) using the keyboard
to control the VCO's.
iii- Repeat step (ii) but with the
keyboard voltage only connected to the
slave VCO(s).
iv. Note the effect of frequency
modulating the slave VCO(s) since this
technique is capable of yielding some very
pleasing timbral effects.
4.8^
It is evident from Figure ^.8.1 that
NEGATIVE HARD SYNCHRONISATION (-
HS) may also result in the generation of
complex waveforms and in many instances
it will be impossible to audibly
discriminate between the two hard
synchronisation techniques. To obtain
negative hard synchronisation, the pulse
output from the master VCO is connected
to an 'inverter' on the SO-5 module (refer
to module description) prior to being
distributed further, if required, by an
attenuator and then connected to the -HS
input on the slave VCO's. It is possible to
use both positive and negative hard
synchronisation simultaneously but this
will require two master VCO's set to
different intervals. The latter arrange-
ment, as well as the use of other sources
of synchronisation pulses, may be
interesting but their complexity is such
that they are unlikely to be rewarding in
the early stages of experimentation,
SOFT SYNCHRONISATION causes prem-
ature reversal of the waveforms from the
slave VCO's with the result that their
oscillation period is an integral multiple
of the pulse period of the master VCO.
Soft synchronisation (SS) requires
negative going pulses (positive pulses
should be avoided entirely) and so may
only be implemented in the same manner
as described for negative hard
synchronisation. For soft synchronis-
ation, however, the attenuator on the
'inverter', or the distribution attenuator,
should initially be set mid way and then
gradually increased, if necessary, until
synchronisation occurs. The nominal
pulse amplitude for SS is ~5 volts
maximum. When setting up for this
harmonic locking effect the master VCO
should be tuned for the correct interval
and the fine control adjusted to an exact
frequency lock, which will be audibly
evident. It will be obvious that turning
the attenuator control on the Processor
anti-clockwise will stop the synchronis-
ation from taking place. We can develop
some simple but effective techniques
from this ability to suddenly take VCO's
in and out of synchronisation. In Figure
^.8.3 a simplified diagram is shown of a
basic patch for +HS and it may easily be
converted for -HS or SS by using the
methods outlined above. The VCA may
be amplitude modulated by another pulse
to take the VCO's in and out of
synchronisation, thus • providing some
unusual effects. Alternatively, an
envelope voltage contour may be applied
to the VCA so that synchronisation only
occurs for part of the contour. The best
VCO
1
VCA
r
r
VCO
2
VCO
3
4&3
type of envelope for the latter is one with
near minimum attack time and high
sustain levels (to effect the sync.) and
thus the VCO's come out of synchronis-
ation as soon as the release phase
commences. It will also be best to have
a fairly short release time. This effect
must, however, be precisely set up in the
overall patch otherwise the results will be
jittery as soon as the key is released. It is
nevertheless a useful approach for
obtaining automatic control of the
synchronisation function and it should be
examined.
It should also be borne in mind that the
VCLFO's modulation frequencies can be
synchronised in the same way as a VCO
and as a final experiment in this section
apply the synchronisation techniques to
the cross modulation patches described
earlier in this manual. It may also be
applied to other patches where multiple
VCO's and VCLFO's are used.
45.1
4.9 FURTHER DYNAMIC CONTROL METHODS
We have already explored several v^ays of
altering the periodic and tiresome nature
of some electronically produced sounds.
In this section we introduce additional
ways in which properties of a sound can
be varied with time and the emphasis will
be on the keyboard to provide the
dynamic control methods.
At this stage, we should take a fresh look
at situations in which use is made of a
VCLFO. The patch of Figure 4.9.1 shows
how the keyboard voltage is also used to
control the frequency of the VCLFO.
This quite simply means that the
modulation frequency of the VCO is
entirely dependent on which key is
depressed, e.g., the higher the note played
on the keyboard the higher the rate of
modulation. The use of this technique
contributes a great deal to creative
synthesis, more than will initially be
apparent. For the first series of
experiments in this section we should
apply the technique to patches and
situations described earlier in the manual,
paying particular attention to the patches
shown in Figures 4.4.2, 4.4.3, 4.4.4, 4.4.7,
4.4.8, 4.5.2, 4.5.3, 4.7.1, 4.7.2, 4.7.3, 4.7.9
and 4.7.10. These experiments should be
repeated using an envelope generator
gated from the keyboard to provide the
control input to the VCLFO. As an
example of a working situation Figure
4.9.2 shows a patch which provides a more
realistic simulation of the tremelo effect
previously mentioned in the manual. No
explanation of the patch should be
necessary since it is merely a combination
of techniques which were introduced
earlier. It should, however, be noted that
the 80-10 VCEG is in the delay mode and
the built-in timer set such that an AD
envelope is generated coincident with the
end of the attack period of the other
envelope generator (ADSR). The AD
envelope from the VCEG is obtained by
setting the sustain level to zero, as
previously described, and the envelope is
also used to vary the frequency of the
VCLFO during the remaining portion of
the note. The reader should also now be
in a position to modify this patch for
himself and so enhance Its effect.
Experiments can also be tried using a
VCO modulating another VCO, e.g., by
patching the keyboard control voltage to
the 'normal' signal VCO and a portion of it
4^.1
SIGNAL SOURCE
WITH TftEATMENT
-* VGA [)
VCLFO
3
rc2
VCEG
KBD.
QATi
I
ADSR
AJd.2
to the modulating VCO via Control Input
2. This will yield somewhat similar
effects to the cross modulation patches
described earlier except that in this case
somewhat strange, though still quite
tuneful, sounds will occur without too
much wandering from the keyboard
related notes.
Since the Voltage Controlled Amplifier (i
of 80-9) can be DC coupled further
methods of dynamic control are available
using this module.
DC
VCLFO M
KBD.
u
ADSR
ADSR
AA3
HI
4A4
VCO [^ VCO -^ VCF -^ VGA i^
'C2 JfM TC2 ^ tl ^
VGA U VGLFO -^ VGA
ADSR
Examine the patch of Figure 4.9.3. The
arrangement is a technique known as
•dynamic depth frequency modulation'.
The title really explains the action. The
amount of VCO modulation present varies
in proportion to the envelope contour.
The technique, may also be used for
modulation of filters, PWM, etc. and
again we should examine some of the
previous patches in the manual and apply
Dynamic Depth Modulation wherever
applicable. To provide a variation, use
can also be made of the keyboard control
voltage to alter modulation levels in a
more subtle way by patching to the VCA
AM input(s). The X-Y Controller may
also be utilised in several situations where
a subtle change in frequency or depth is
required, or indeed the initial pulse width
of the VCO may be set using the keyboard
control voltage such that the higher the
note the wider the pulse. To provide a
final example, a patch is shown in Figure
4.9.4 which may be used for the creation
4.9^
of a sound which is best described as a
space war weapon! This is simply the
extension of the previous patch. The
release time of ADSR 1 must be shorter
than the release time of ADSR 2. The
manual gate may be simply performed by
pressing the two manual gating buttons on
the ADSR units simultaneously or by
patching in the keyboard gate and
pressing the key. The patch will result in
modulation of the signal mostly during the
period when the gate is 'on'.
4.10.1
4.10 THE USE OF NOISE
IN ELECTRONIC MUSIC
We have now reached the point in
synthesis where we need to create sounds
other than musical ones and one of the
most important modules for the
production of such sounds is the Noise
Generator. The sounds of electronically
created wind and howling gales, utilising a
Noise Generator, are to be found on
hundreds, if not thousands, of albums and
therefore we will attempt to progress a
stage further and use this module for
other applications. We must first learn,
however, to create the basic sounds.
The effect of wind may be created simply
by patching the white noise output to the
input of a voltage controlled low pass
filter. Simple manual control of the cut-
off frequency of the filter is all that is
required to generate a wind sound.
Careful setting of the resonance control
on the 80-6L filter can produce a 'whistle'
in the audio output which enhances the
effect. The best setting of this latter
control is where the filter is just about to
break into oscillation. Substituting white
noise with pink noise will create a sound
of deeper intensity, more like the sound
of the sea. We may experiment with
these basic patches by adding the
keyboard control voltage to the filter's
control input and so effectively 'play' the
wind or sea. Adding an envelope
generator to the filter's control input
while still using the keyboard voltage can,
if careful attention is paid to envelope
shape, result in a breaking wave and this
may be brought in when required by
manually gating the envelope generator.
Figure 4.10.1 shows a simple patch which
may be used to simulate gunfire.
Ricochet may be added by turning up the
resonance control until the filter actually
goes into oscillation during the sweep
time. Longer envelopes may be used to
simulate explosions. Now try the
following refinement. Patch the low
noise output to the external resonance
control of the filter. The result is a very
powerful deep rooted modulation that
greatly enhances the rumbling of the
explosion. You should also examine the
use of this low noise modulation technique
with some of the cross modulation
patches described earlier in the manual
since some very impressive effects are
possible. Figure 4.10.2 is an extension of
iWHtTE
I NOISE
•■ VCF-L — I
prNK
NOISE
12-,
LIHw
ADSR
LOW
NOISE
VGA
PINK
NOISE
Qi
4.10.1
4.10.2
VCf-L
31
ADSR }•* MAN.
GATE
the previous patch which allows the low
noise to be faded out of the filter as the
envelope progresses. The manual gating
buttons should be used for this
application.
Next experiment with the direct injection
of noise to the VCF control input, via the
attenuator, which will result in a strange
and often 'nasty electric edge' to the
sound. It will be noticed that when the
low noise source is used in this way a very
pleasing random filtering effect will be
given to the sound, especially with high
levels of attenuation so as to avoid
saturation of the control input. The
reader should also revert to some of the
earlier patches in the, manual and
experiment with the injection of noise
into a control input of a VCO. Now try
treating noise through various types of
filter, including the lag processor on the
80-5 module, and, in particular, the 80-6P
phase shift filter if available. Also mix
in some conventional audio input, i.e., a
normal synthesised sound. Make a note of
the results obtained for future reference.
Next revert to the introduction of noise
to the control input of the filter when the
signal input of the filter also comes from
the noise generator. The use of an 80-6P
filter with a pink noise input and low
noise for control of frequency and/or
resonance will produce some interesting
results.
Automatic sweeping of filters using the
contours from VCLFO's or ADSR's may be
used effectively with noise inputs. Figure
4.10.3 shows an initial patch for the
automatic creation of surf sounds. As an
exercise, try perfecting the patch using
some of the techniques discussed earlier.
The following suggestions should help:-
i. Interpose a lag processor between
the VCLFO and the VCM to modify the
4.10.2
NOISE
ATT.
SI
VCM
77 "
VCLFO
r* VCF
NOISE ^* VCF Ir VCM
tF=f
!-»■ VCF -i
ITT
a nc3
ADSR
ADSR
4.10.3
4.10.4
ADSR
transition in sound v^hich occurs when the
sawtooth wave reaches its peak
amplitude.
ii. Add the low frequency noise source
to Signal Input 2 of the VCM and adjust
the sound level of this signal manually.
The use of low noise provides a
background rumble, often heard with
strong seas, and it is necessary to obtain a
good balance between the intensity of the
two noise sources now being used.
iii. Instead of a sawtooth waveform
use an envelope voltage to control the
sound contour of the noise sources. The
envelope generator may be gated
manually or from the keyboard. The
time constants should be set to provide a
long attack, fast decay and a fairly long
release time so that the latter simulates
the effect of the receeding waves. A
medium sustain level should be used. The
conventional envelope gives a fast initial
build up of sound but slows as the attack
proceeds. An. improved contour would be
one with a slow build up and rising to a
crescendo as the wave breaks. This type
of contour may be obtained using the 80-
10 VCEG in the following manner:
Connect the output of the 80-10 to
an attentuator on the 80-5 module
with the potentiometer fully
clockwise since it is to be used
solely for distribution purposes.
b) Take one output from the
attentuator to the '-' input on
Attentuator 1 of the 80-10 and
from 'Al' to the external attack
control 'A'.
c) Adjust Attentuator Al and the
attack control potentiometer to
produce a concave attack response
of desired shape and time.
d) If waves of greater ferocity are to
be simulated then the white noise
may be high pass filtered. This
filter should be directly after the
white noise source and not on the
a)
e)
final sound since the latter
arrangement would result in the
low frequency noise being filtered
out.
Try the patch substituting the
white noise with pink noise.
The above demonstrates the method of
approaching sound synthesis. That is,
building up a picture of the sound in your
mind which in the above case is one of
gradually increasing intensity of sound as
the wave approaches and the fairly sharp
decay as the wave crashes on the shore.
Thus initial sawtooth shaping of white
noise can give a reasonable imitation but
this may be greatly improved by slight
modification to both contour and nature
of the sound source.
A further exercise is the development of
the patch shown in Figure 4.10.4. Also
experiment with variations in the
technique to produce a series of sound
effects which may prove useful at a
future date.
Noise is also very useful for creating
percussive effects and may be tried using
noise as the signal in the patch of Figure
4.7.6. By using synchronisation
techniques to ensure exact timing
relationships this patch may be expanded
into dual, triple, quad, etc. 'voice' and the
frequency of one VCLFO may be set to
produce timing pulses which are integral
multiples of another, thereby producing a
basic rhythm. The example in Figure
4.10.5 shows a two voice percussion
patch. Synced VCLFO's are used and
indicated by the markings on the diagram
and, for clarity, it is not normal to show
the type of synchronisation used. A
moments thought, however, will reveal
that Soft Synchronisation (SS) will work
best here. Refer back to Section 4.8 if in
doubt.
.«.!
r*
VCF
—
.. 1
■*
VCm p
■
ADSR
A
r-H
SYNC
VCLFO 1
a
Igat
.i L
^
C>
VCLFO
n
GAT
ADSR
— p-
A
Ef '
4.10 .5
VCF 1
r-J-n
"
VCA L
J
The noise source may, of course, be
replaced by two VCO's to provide 'pitched'
percussion using the above patch or even
one VCO and a noise generator. Conduct
some experiments with percussion
4.10.3
techniques and extend your ideas into
multi-voiced patches if resources allow.
Noise may also be used to simulate the
sounds of certain engines or other pieces
of machinery and the resultant effect
may be used on its own or added to
another synthesised sound to provide a
background for the final mix. Automatic
control of the noise generator is usually
required for such applications and for
purely 'mechanical' sythesis the use of
ADSR units may often be omitted. The
example in Figure 4.10.6 uses a VCLFO to
'open' and 'close' a VCA to provide
amplitude modulation of the noise signal.
Use is also made of the ability to achieve
a 0% duty cycle (zero output) of the pulse
waveforms on the VCO's thus causing the
sound to be 'off for part of the VCLFO
waveform cycle. The three sound
sources may be mixed in any proportions
in the VCM, after which further
treatment may be applied.
frequencies and pink noise equal power
per octave. This fact alone may be used
to create totally unpredictable events and
some of these are applied in the next
section dealing with the Sample & Hold
section of the 80-12 Noise Generator.'
vco
PWM
PWM
VCLFO
51^
vco
VCM
r'
M
INCISE
-
VCA
vco
vco -t-[)
vco OR
VCLFO
\!'
4.10.6
4.10.7
It will be observed that certain patches,
at some settings of the controls, will
produce an output which may be referred
to as 'noise' in spite of the fact that
periodic events are occuring within the
complex waveform of the output. Such
patches may be used in place of a noise
generator and are quite effective for
some sounds, for example, 'mechanical'
type of sounds. It is generally advisable
to use as few modules as possible to
create this 'periodic noise' since they may
be required for some subsequent
treatment of the effect. A band pass
filter, finely tuned, works well in this
application since often only certain parts
of the sound are required at the output
and may provide the entire effect that is
sought.
As a starting point for other patches
which will create a complex output try
the patch illustrated in Figure 4.10.7 with
high settings of both frequency and
modulation depth.
The nature of a true noise signal consists
of a mixture of all audio frequencies with
white noise having equal power at all
4.11.1
4.11 APPLICATIONS OF SAMPLE & HOLD
In the previous section we examined the
use of a low frequency noise source to
provide some random events but these had
a severe limitation, namely, that they
could only be altered in terms of level.
The principal use of the 80-12 Sample &
Hold unit is the production of random
voltages which, to a certain degree, may
be manipulated by the user so that, for
example, their speed may be altered at
will. By having a module which produces
random voltages at useful rates we can
utilise the facility to provide a means for
the synthesiser to play itself
automatically.
The basic patch is shown in Figure 4.11.1.
The low noise from the same module is
connected to the S&H input, the output
from the 5<5cH to a VCO and an
appropriate waveform from the VCO
direct to an amplifier. A series of
discrete pitches which are random in
frequency are produced and their tempo is
determined by the setting of the internal
clock. The mid frequency of the output
is set by the coarse control on the VCO
while the total range of pitch is
adjustable with the attenuator on the S&H
input or the Control Input 2 attenuator on
the VCO, or both. The effect is
psychologically very powerful and in a
more simplified manner this is generally
the only application of Sample & Hold in a
small synthesiser and thus has^ become a
well-worn product of electronic music.
p/>^
LOW
S+H
VCO
NOISE
L_
^
]0
4.11.1
The random voltage from the Sample &
Hold unit may also be used to control the
frequency of a filter. The 80-6L low pass
filter works best here especially when the
initial frequency (set by the coarse
control of the VCF) is set to almost
totally filter out the audio input.
Furthermore, by carefully increasing the
resonance the filter will begin to oscillate
and the result is a series of random sine
wave tones accompanying the modulated
audio frequency. If the step voltage
change from the S&H is too obtrusive for
a particular application then the random
voltages from the S<ScH output may be
taken via the 80-5 lag processor to
'soften' the edges of the sharp transitions
without interfering too much with the
overall effect. The latter is similar to a
portamento control.
We should now experiment with various
applications of random voltages and the
following provides some guidelines:
i. Try the pink and white noises as
the material for sampling.
ii. Apply filtering to the noise sources
prior to sampling, with emphasis on the
band pass filter which will reduce the
bandwidth of the source.
iii. Introduce random frequency
control to VCLFO's with particular
attention being paid to patches that
utilise some form of dynamic control.
Random, or random frequency, pulse
width modulation is another technique
which should be fully explored.
In a modular synthesiser, however, the
scope of application of the S&H unit is
virtually unlimited. The S&H input will
accept signals from virtually any source
and these may be pre-recorded material,
suitably amplified by the 80-13 External
Input module. Generally, however, the
result of sampling most external material
is still to produce a series of random
pitches. A more useful signal source for
the S&H is the sawtooth waveform from a
VCLFO. The latter will produce discrete
pitches which are gradually rising in scale
until the peak of the sawtooth is reached
after which the stepwise scale repeats,
although not necessarily with identical
frequencies. The illustration of Figure
4.11.2 makes the general principle clear.
I I
i i I I 1
SAWTOOTH INPUT
S&H OUTPUT
-" SAMPLING (CLOCK) PULSES
4.11.2
4.11.2
At low input frequency several notes may
be obtained during one cycle of the
sawtooth waveform and by adjustment of
the S&H input (or Control Input 2 on the
VCO) the steps may be of widely varying
frequency intervals. As the frequency of
the sawtooth is increased one begins to
obtain arpeggiation effects and then more
complex sound patterns are created as the
frequency of the sawtooth exceeds that of
the clock. Note that at low frequencies
the 'scale' of pitches may be made to rise
in an exponential manner by interposing
the 80-5 lag processor between the
VCLFO and the S&H input. Another
effect is to modulate the VCLFO which is
generating the sawtooth so as to produce
changing patterns of 'scales' and
'arpeggios'. This may even be extended a
stage further by taking a proportion of
the random voltage back to the control
input of the VCLFO used to modulate the
VCLFO being sampled. A note of the
effect should be made.
The S&H unit may also be used with an
external clock, for example, the pulse
waveform from a VCLFO. To obtain the
discrete frequencies referred to above it
is necessary to set the duty cycle of the
pulse output near to its minimum. This
limitation on pulse width has been
purposely designed into the module in
order to increase the range of application.
For example, when sampling a sawtooth
waveform if the duty cycle of the VCLFO
clock pulse is increased then a gliding
(portamento) effect between notes will be
obtained and the degree of this effect is
related to the duty cycle, the VCLFO
sampling frequency (clock rate) and the
frequency of the sawtooth being sampled.
To increase variety the pulse width may
be automatically modulated so as to
produce a variation from 'clean* notes to
notes which are slewed to various
degrees. Increasing the sampling time by
increasing the pulse duty cycle can also
be effective when sampling other signal
sources.
Thus far the effects have been at a
uniform tempo using either the internal
clock or a LFO. Since, however, the LFO
is voltage controlled this restriction may
be avoided. A simple means of achieving
variation in tempo is to modulate the
clock VCLFO with another VCLFO which
will cause the clock to speed up and slow
down in a controlled manner.
Alternatively the output from the S&H
module may be connected to the Control
Input 2 of both the VCO (to generate the
pitches) and the VCLFO (to vary the clock
speed). In this patch a high output from
the S&H will effect both a high pitch and
at the same time speed up the VCLFO so
that it takes the next sample faster.' The
effect is therefore more musical since the
higher the note the shorter the duration
of the note. Clearly one may also obtain
the opposite effect by interposing an 80-5
'subtracter' between the S&H output and
the VCO and/or the VCLFO control input.
In the above patches the S&H directly
drives a VCO which is connected to an
amplifier without any intermediate
modification to timbre or shape. Having
created sounds using the keyboard
programme described earlier the user may
wish the Sample & Hold unit to generate
some of these sounds automatically. To
accomplish this the arrangement would be
similar to Figure 4.1.1 but with the
keyboard replaced by the S&H unit. The
main shortcoming, however, is a lack of
control over the duration of the gate
pulses which determine the sustain period
of a note. The clock output from the
S&H unit, or the pulse from a VCLFO,
may be used as an independent trigger for
the 80-10 envelope shaper but they should
not be used for gating this module without
first attenuating them to +5V (unless the
80-10 has been modified to accept higher
gate voltages). Either type of clock may,
LOW
NOISE
S+H
SKS
EX
T'
CK
T
VCO
-J
VCF-L
-^
vc.
>
t"
f-^
r
AOSR
1
ADSR
2
ADSR
3
VCLFO
n
:
t
t
G*Tt
4.11.3
however, be used to gate the 80-8
envelope generator directly. Owing to
the short duration of the clock pulses in
normal operation they are only suitable
for percussive (AD) type envelopes. This
limitation is overcome with the patch of
Figure 4.11.3 in which the pulse width
from the VCLFO may be varied over its
range to generate the gate time for the
80-8 ADSR's 2 and 3. The controls on the
80-8 ADSR 1 are all set to minimum so as
to generate a sharp pulse for sampling
which is independent of the pulse width of
the VCLFO. In this arrangement the
output from the VCO may be treated to
generate a variety of sounds as described
in Section 4.2. It will be evident that
practically all of the techniques so far
described for the S&H and for the
keyboard generated sounds may be applied
to this patch. A further possibility is to
4.11.3
use the S&H output for PWM of the
VCLFO clock.
In the patch of Figure 4.11.3 the VCLFO
may be replaced by the keyboard. In this
situation the pitch is still determined
independently from the signal being
sampled but the tempo and duration of
notes is controlled by the player. Various
combinations of keyboard and S&H are
worth exploring and a useful combination
is shown in Figure 4.11.4 in which the
keyboard gate output is used to control
the envelope generators and the keyboard
control voltage is employed to control the
frequency of a second VCO. This latter
VCO is used as a master oscillator
providing positive hard synchronisation
pulses to the VCO whose frequency is
being controlled by the S&H. Adjustment
of the various control inputs will result in
a situation where just pressing the same
key will generate a series of tones whose
pitch is unpredictable but are
nevertheless musically quite pleasing.
This musical effect is enhanced by playing
the keyboard in the conventional manner.
The best results are obtained when the
VCF is tracking the VCO and for this
arrangement the S&H output goes direct
to Control Input. 1 on both VCO and VCF
while the pitch spread is set by the
attentuator on the S&H input. We can
simplify the technique illustrated to
provide a method of deriving a random
voltage to accompany each note played.
Figure 4.11.5 shows a typical keyboard
patch which provides a random
accompaniment to modulate the VCF and
VCLFO.
Methods of limiting the range of the S&H
output should now be familiar and these
should be applied to the patch of Figure
4.11.5. The basic technique of random
vibrato can easily be deduced and the
method can be used for a variety of
situations and not just with the dynamic
depth frequency modulation shown.
One of the most interesting patches for
the S&H is illustrated in Figure 4.11.6
which allows the creation of a wide
variety of rhythmic patterns of a quality
which is normally only available using a
sequencer. Start by manually adjusting
the frequency of VCO 1 slightly higher
than VCO 2. Adjust the 5&H clock for a
rhythm tempo and adjust both input level
to S(5cH and Control Input 2 on VCO 2 to
give a good dynamic range while avoiding
very low and very high frequencies. All
of these adjustments have a major effect
on the resultant sound. What happens in
this patch is that sometimes the S&H
output is taking the frequency of VCO 2
above that of VCO 1 and producing a
variety of timbral effects through
synchronisation whereas at other times
yCO 2 frequency is lower than that of
VCO 1 and so the output is determined by
VCO 1. The resultant sound is a series of
changing frequencies but with some of the
notes also changing significantly in
timbre. The sound may be made more
rhythmic by replacing the low noise signal
with a sawtooth waveform, with or
without slewing by the lag processor.
With the latter the result is an 'easy-to-
listen-to' rhythm which may be used as a
background track. Because of the
interesting nature of the output this
apparently simple patch can consume
several hours in learning how to obtain a
variety of rhythms. This patch may
obviously be extended, e.g., modulation of
the sawtooth and treatment of the VCO 2
output as described earlier.
LOW
NOISE
ADSR
1
S+H
clock]
Ici
VCO
2
'^fol*! VCF U VGA |[N
ADSR
T
KBD.
KBD. =J
ADSR
ADSR
ADSR
NOISE
VCLFO
IxtI L
> CKf
* S+H
VLIN
VGA -1
VCO M VCF [♦
VGA
10
)—"*• ADSR
LOW
NOrSE
^J
Tl
C2
vcp
+HS
y/C-n
^
11^
4.11.6
4.11.4
4.11 .5
4.114
A more advanced technique of random
voltage applications is shown in Figure
4.11.7. Again with this patch a great
deal of patience is required to obtain
useful results. The Sample & Hold is
clocked by the keyboard in the manner
previously described. The output from
the S&H must be restricted so that some
of the random output voltages will fall
below the minimum voltage required to
gate the ADSR envelope shapes. Thus
attenuators have been placed in the lines,
as illustrated, to facilitate different
voltages being required to gate each
ADSR. As may be seen from the patch, a
normal note will first be sounded followed
by either VCO's 2, 3 or 4 and each
different note will yield a different
combination. Each VCO may, of course,
be treated by filtering and so on prior to
entering the VCM and this can result in
some very interesting and unusual effects.
LOW
S'l-H
NOISE
cSdc
KBD.
P.fTE^
AOSR
^ ATT. S4BE» ADSR
ATT.
AOSR
!-♦ ATT. fifil^ ADSR
VCO
4
VCO
3
VCO
2
-IS-
VCO
1
I:
VCM
4.11.7
This basic idea may be used without the
random influence by using the keyboard
via a 'subtracter' (to generate higher
voltage levels) and the voltage from the
'subtracter' used in the same way as the
voltage from the S&H in the patch of
Figure 4.11.6. The normal (non-
subtracted) keyboard voltage would be
used to control the VCO's in the same
way. This patch will result in a lesser
amount of VCO's being brought into use
the higher the note on the keyboard. For
the best results with this technique one
should avoid playing slow pieces for the
simple reason that somewhere on the
keyboard two notes a semitone apart will
be the difference between one or two
VCO's being sounded. The latter occurs
for three different points on the
keyboard. For slow pieces the technique
of using four different ADSR/VCA
combinations (if four VCO's are being
used) is much simpler to set up since the
attack times only need to be adjusted for
the fading in of different VCO's (or
voices). The VCM may even be used to
perform the task of the four VCA's thus
offering a saving on module count. The
ADSR units are gated from the keyboard
in the normal manner.
Figure 4.11.8 shows the S&H in a 'working'
patch. Note the use of the clock output
of the S&H. The clock output, being a
narrow pulse, may be used as a trigger
input for DIGISOUND 80 envelope
generators but in this application it is
being used to 'blip' the filter each time a
sample is taken. The VCO is set to
maximum frequency and the sine wave
output is utilised. Normally this will not
filter well but in this patch the frequency
zr
-«
D
'■c,(
[CI C2
CLOCK
OUT]
LOW
NOISE
PINK
NOISE
-*
S+H
4.11.8
int.clx3>
modulation via the Control Input 2 adds
sufficient harmonics for the desired
effect to be created. Note also that the
settings of the control potentiometers
have been included and are indicated by a
number within a circle. The resonance
control of the filter should be set to the
point where oscillation is about to take
place. The initial frequency of the VCF
is set to pass only very low audio
frequencies. This patch is reminiscent of
a 'water-drop' effect at low to moderate
sampling rates. Increasing the frequency
of the clock and also the cut off
frequency of the VCF will give an effect
similar to running water.
4.12.1
4.12 RING MODULATION
The ring modulator may be used with two
input frequencies, X and Y, and its output
is the sum and difference of these two
frequencies, that is, X + Y and X - Y.
With sinewaves the resultant outputs are
well defined. When one or both of the
inputs is, however, a waveform with a
high harmonic content then the resultant
output becomes extremely complex. The
first experiment is illustrated in Figure
4.12.1 in which the ring modulator module
is denoted by the multiplication sign in
the box. Try all combinations of
waveform outputs from the VCO's
including the +5V sine and triangle
outputs. Next try controlling one of the
VCO's from the keyboard after setting the
rest of the patch up in the same way as
Figure 4.1.1. The result of these
experiments will normally be a series of
heavily modulated sounds ranging from
deep gong effects to less harsh chimes. It
will also be noted that in many instances
only a small change in frequency will have
a pronounced effect on the quality of the
sound.
Experiments should progress in a logical
manner and Figures 4.12.2 to 4.12.6
illustrate some basic variations which
should be further developed. Note that
Figure 4,12.3 allows for altering the
harmonic structures, hence the ring
modulation effect, of the input to the ring
modulator. The patch of Figure 4.12.6
will hold certain frequency ratios
constant and a further development
(phase locked loop) is shown in Figure
4.12.7.
vco
vco
r
x]0
VCO
i
jVCLFO
VCLFO
4.12.1
h
aS]!)
VCO
4.12.2
VCO
VCLFO
onAbsn
vco -*
^0
4.12.3
4.12.4
VCO -I
AOSR
VCO
VCO
^SO
vco ->
KBD
VCO
f
><\t\>
4.12.5
4.12.6
A patch which has been used in a working
situation is shown in Figure 4.12.8 and
this illustrates the extent to which patch
development may be taken.
Returning once more to the dual
percussion patch of Figure 4.1Q.5 we now
show a variation (Figure 4.12.9) which
may be used to provide a w'de range of
ring modulation effects. Note that the
ADSR's should be adjusted so that there
are always two signals being applied to
the ring modulator. This patch, in
-J X -* VCF-L -|-|>
-^ORJl
vco
'0 tea.
4.12.7
VCLFO
vco
VCLFO
C2'' VPWM
%
VCLFO
CI
KBD.
VCO
J^ %
vco TTVCF
LOW
NOISE
X
VCM
4 J' A
C2
VCF
VCA
.«!.
SUBT.
ADSR
ADSR
4.12.8
4.12.2
VCO
VCLFO
""lL VCLFO
n
VCO
VCF
1-^ VGA
icxf
^-^ ADSR
!^ AOSR
^XO
VGA -*
RANDOM
4.12.9
common with many others, requires
patience in setting the controls in order
to achieve the desired effect. There are,
of course, many simpler ways of obtaining
a percussive type sound with different
timbral qualities on each 'beat'. The
advantage of the method illustrated here
is that the random steps may be
synchronised to one of the LFO's so that
the timbre changes will be initiated after
a certain number of 'beats'. The patch
yields some unusual results.
4.13.1
4.13 THE EXTERNAL INPUT
Rhythmic patterns from the synthesiser
for use as background tracks are, as
already inferred, normally obtained by
using a sequencer which controls both the
tempo and VCO frequency. In the
absence of a sequencer, or the
ALPHADAC 16 computer controller, one
may use the Sample & Hold module as
described in Section 4.11 or the External
Input module. The results from the
latter modules are, however, very limited
in comparison with the aforementioned
controllers. A patch using the 80-13
module is shown in Figure 4.13.1. In this
patch the VCLFO is amplitude modulating
the output from a VCO in the 80-4 VCM
to form a well defined series of peak
voltages which are subsequently detected
by the GATE/TRIGGER input of the 80-13
External Input module. The derived gate
pulses are connected to an 80-10 VCEG
for shaping the sound output from VCO 2
in the VCA. VCO 2 could be a sawtooth
output from VCO 1 but this reduces
flexibility.
VCO
nsi
\ir
G/T
DETECT
GATE
A05R
vc»
^
■
Ci
\
VCLFO
VCO
VCA
,
'
4.13.1
^
The range of rhythms may be extended by
modulating the VCLFO with another
VCLFO, or a VCO at low frequency. Tone
shaping may be accomplished by
interposing a low pass VCF between VCO
2 and the VCA and the centre frequency
of the latter varied with a triangular
waveform from the VCLFO. This latter
VCLFO may be the modulating VCLFO
illustrated but again using the same
module for two purposes obviously
imposes limitations on the variety of
sound obtainable. The 80-10 VCEG can
make use of both the gate and trigger
outputs when its function switch is in the
DELAY mode but the effectiveness
depends on the degree and variety of
amplitude variation going . to the
GATE/TRIGGER extractor of the 80-13
module. Furthermore one has to ensure
that the gate and trigger pulses are
adequate for driving the VCEG and this is
indicated by the status LED's on the 80-
10. If necessary the input to the 80-13
should be adjusted via the gain control on
the VCM.
[PRE-
■ .
G/T
1
AMP
'
DETECTl
GATE
\
1
ADSR
ADSR
\ 1
VCF
—
VCA
4.13.2
To use the 80-13 for interfacing with
external equipment one should use a patch
of the type shown in Figure 4.13.2. This
patch works best when the input material
has a series of well defined peaks and it
allows for filtering and envelope shaping
of the input signal, or other voltage
controlled modifications that may be
required. When the peaks are difficult to
extract from the source material then a
modification to the patch will be
necessary. Figure 4.13.3 shows one
alternative which takes the external input
through the envelope follower and so
smooths out the waveforms amplitude
variations. The output from the envelope
follower may be used to directly control a
VCA (linear input) since the DC voltage
obtained is directly proportional to the
amplitude of the external input signal.
Depending on the nature of the input
material there may be a residual voltage
mixed with the envelope which prevents
the VCA from completely cutting off but
this is easily remedied by partially
•burying' the envelope, as described in the
80-9 VCA description notes. The output
of the envelope follower also goes to the
GATE/TRIGGER detector in order that an
ADSR contour may be derived, as in
Figure 4.13.2, which in turn is used to
control various voltage controlled
functions. In the patch of Figure 4.13.3
it is shown controlling a VCF.
^ DQC— - - 1 ENV 1 - -
ADSR
ie:
G/T
iDETECT]
VCF
VCA
4.13.3
Experiments should be made using
different types of external sources such
as voice, guitar, pre-recorded music, etc.
and the various types of signal
4.13.2
modification applied. With some
external signals it is worth filtering them
after the pre-amplification stage so as to
facilitate subsequent extraction of their
envelopes or their gate and trigger peaks.
Finally, if one simply requires to modify
the external signal after the pre-
amplification stage using the facilities
provided with the DIGISOUND 80 then the
following guidelines should be of
assistance. Voice signals respond well to
ring modulation and phasing and also
treatment by random filtering. For ring
modulation the voice is used as one input
while a VCO provides the modulating
signal. The resultant sound can be
'Dalek' in character depending on the VCO
frequency used. A more characteristic
'alien' sound may be obtained by
subsequently 100% amplitude modulating
the signal in a VCA. Guitar signals
respond well to additional filtering and
amplitude modulation and again the ring
modulator can give rise to some strange
effects which are perhaps best mixed with
some of the original signal. By this stage
in the manual you should, however, be
well acquainted with the facilities offered
by the various modules and thus in a
position to decide for yourself the best
treatment to apply to an external signal
in order to obtain the desired effect.